129 results about "Homogeneous transformation matrix" patented technology

The present invention relates to the field of robotic arm teleoperation technology, and specifically to a construction method and system for real-time teleoperation of a dual arm and hand robot based on vision guidance, which includes setting up a first experimental platform, acquiring pose information of the dual arm and hand robot and a homogeneous transformation matrix of an end joint and transforming a first coordinate system; setting up a second experimental platform; performing calibration by a plurality of cameras and transforming a second coordinate system; obtaining spatial transformation pose information of the end joint of the robotic arm under a tracking state, reading finger pose data, and performing real-time following. The real-time teleoperation system of the present disclosure includes a control system host, an operation glove, an optical locator, a positioning coordinate plate, a marker ball tool, and a dual arm and hand robot. The present disclosure introduces a method of identifying the marker ball tool by the optical locator into the teleoperation control of the dual arm and hand robot, which is more accurate and faster. Meanwhile, the method of remotely controlling the robotic arm by an operator wearing the operation gloves can provide better human-machine interaction function.

The invention discloses a dynamic trajectory control method for an SCARA (selective compliance assembly robot arm) robot. According to the kinematics model, a homogeneous transformation matrix is used for analyzing the coordinate transformation relation between each joint and the connecting rod, and an initial pose and a target pose are determined; establishing a kinetic model based on a Lagrange equation, and determining kinetic parameters needing to be identified; performing parameter identification by adopting a random weight particle swarm optimization algorithm; according to an identification result, a sliding mode robust item RBF neural network adaptive controller is constructed, real-time track control is achieved, and the end manipulator is made to move to a target pose along a preset dynamic track; a real-time pose is detected through a sensor and compared with a target pose, and arrival is judged when the error is smaller than a threshold value. Model precision and identification efficiency are improved through modeling and efficient parameter identification, the composite controller effectively compensates uncertainty, disturbance and nonlinear factors of the model, closed-loop control and online adjustment are achieved, and trajectory tracking precision, system stability and dynamic response speed are improved.

The invention belongs to the crossing field of computer vision and three-dimensional reconstruction, and particularly relates to a large-scale three-dimensional scene registration method. A 3D Gaussian sputtering large-scale scene registration method fusing GPS prior information comprises the steps that S1, an unmanned aerial vehicle carrying a GPS receiver is used for conducting data collection on a registration area, GPS / image information collected by the unmanned aerial vehicle is obtained, and sparse point clouds are produced through feature extraction and matching and camera pose estimation by means of spatial position prior information provided by a GPS; initially reconstructing the SFM model; step S2, model alignment: adopting ENU coordinate system conversion and ENU homogeneous transformation matrix calculation to establish initial reconstruction SFM model alignment so as to reduce a registration search space; s3, efficiently reconstructing a dense point cloud model in combination with a 3DGS technology; and S4, finally, high-precision alignment is realized through two-stage optimization of GPS coarse registration and fine registration. According to the method, the problems of low calculation efficiency, low convergence speed, limited precision and the like in large-scale three-dimensional scene registration are solved.

The invention discloses a calibration method, device and equipment for an external actuator of a robot, and the method comprises the steps: enabling the tail end of a mechanical arm with a calibration plate to be in contact with the actuator in different poses, and obtaining two different circular mark points and a preset transformation matrix; calculating a three-dimensional point cloud coordinate of the circular mark point based on a mapping relation between a pixel coordinate and a point cloud coordinate in the structured light system; calculating a pose transformation matrix between two poses at the tail end of the mechanical arm according to a preset transformation matrix; according to the three-dimensional point cloud coordinates, the pose transformation matrix and a preset hand-eye matrix, the actuator coordinates of the tail end of the actuator under the mechanical arm base coordinate system are calculated; direction vectors of the tail ends of the actuators are calculated through the coordinates of the two different actuators; and calculating a homogeneous transformation matrix calibrated and aligned at the tail end of the actuator according to the direction vector to obtain a calibration result. The technical problem that the actual calibration scene requirement is difficult to meet due to the problems of low precision, complex measurement process, high cost and the like in the prior art can be solved.

The invention relates to a spraying robot spray gun teaching method based on a double-event camera. The method comprises the following steps: calibrating internal and external parameters of two event cameras; the spray gun is provided with at least four mark points, and the rigid offset from the mark points to the center point of a muzzle is obtained; calibrating a tool coordinate system T of the center point of the industrial robot spray gun muzzle; establishing a homogeneous transformation matrix from the world coordinate system W to the industrial robot base coordinate system B; teaching: obtaining event flows of left and right event cameras of the spray gun and time flows of the spray gun spraying paint discharge amount; according to the event flow, the pose of the muzzle center point of the spray gun under the world coordinate system W is obtained and converted into the pose of the muzzle center point under the industrial robot base coordinate system B; and a control program of the industrial robot is generated according to the time flow of the spraying paint discharge amount and the event flow of the muzzle center point pose. According to the method, the motion trail is reconstructed through the double-event camera, the control program of teaching is generated in combination with the spraying time flow, and high-precision reproduction of the spraying teaching process is achieved.

The invention discloses an underwater mechanical arm shake compensation control method for a suspension operation scene. The method comprises the steps that a sensing positioning system is installed at the rigid connecting position of an ROV platform and an underwater mechanical arm, and a positioning mark is arranged at a suspension operation target point; the tail end of the mechanical arm is controlled to move towards a target point, an initial homogeneous transformation matrix of the mechanical arm is established according to the initial frame sensing positioning data, and then the position point of the tail end of the mechanical arm is obtained to serve as an expected value; establishing a current homogeneous transformation matrix of the shaking base according to real-time sensing positioning data, establishing a homogeneous transformation matrix of the shaking base relative to a world coordinate system in combination with an expected value of a tail end position point, and obtaining a quasi-compensation angle value in real time by using a shaking compensation algorithm; and after preprocessing, inputting into a joint controller to carry out real-time shake compensation control on each joint. According to the method, all joints of the mechanical arm are optimized and compensated through the shake compensation algorithm, the stability and controllability of tail end operation of the mechanical arm can be ensured, and therefore the operation safety is improved.

The invention discloses a compensation method for a platform shake error during installation of a transformer substation wallboard, and relates to the field of wallboard installation. The technical problem that the installation position of a wallboard and a buckle deviates due to shaking of a platform during aerial work of a crane is solved. The method comprises the following steps: establishing a base coordinate system; installation data under one-time standard installation is obtained, so that the standard mechanical arm pose during wallboard installation is obtained; the three-dimensional coordinates of the buckle at the actual wallboard installation moment are obtained in real time, and the pose of the compensation mechanical arm is calculated through a homogeneous transformation matrix; and the compensation mechanical arm pose is converted into a mechanical arm control instruction, and the mechanical arm is driven to complete dynamic error compensation installation. According to the technical scheme, errors caused by shaking of the platform are effectively made up, the installation precision is improved, the position of the buckle is recognized in real time through the binocular camera, the position and posture of the mechanical arm at the current moment are calculated in combination with the coordinate transformation matrix, and therefore the movement track of the mechanical arm is dynamically adjusted, and offset errors caused by shaking of the platform in the air are made up.

The invention discloses a remote humanoid mapping control method based on a virtual reality motion capture mechanical arm. A motion data acquisition module transmits a matrix of a head coordinate system relative to a world coordinate system and a homogeneous transformation matrix to a coordinate system conversion module, establishes a base coordinate system and a tail end coordinate system, and establishes a kinematic chain conversion model; extreme position data of forearm extension of a user is collected, a proportionality coefficient is calculated, and a motion space boundary of the mechanical arm is output; pose normalization processing is carried out to form a discretized mechanical arm tail end pose sequence, and then the spatial relation between the mechanical arm sphere model and the obstacle is analyzed; generating a collision-free track by adopting a parallelization gradient optimization algorithm; the trajectory optimization module realizes trajectory smoothness and time optimization by constructing a multi-objective optimization function; and a mechanical arm motion simulation system is constructed, the tail end track precision is verified, and the controller drives the mechanical arm to achieve high-fidelity reproduction of user actions according to the track. According to the invention, high-precision and smooth-path motion migration of the mechanical arm is realized.

The invention is suitable for the technical field of machine tool precision design, and provides a five-axis numerical control machine tool complete machine space error sensitivity analysis method, which comprises the following steps: abstracting a numerical control machine tool into a multi-body system based on a multi-body system theory, describing the incidence relation between the structure of the machine tool and each individual by using a topological structure chart and a low-sequence body array list, and calculating the overall space error sensitivity of the five-axis numerical control machine tool. The geometric error of the numerical control machine tool is analyzed, a generalized coordinate system is established, the position relation is expressed by using the characteristic matrix between adjacent bodies, the mutual relation between multiple body systems is expressed by using a homogeneous transformation matrix, and a space error model of the machine tool is established; and in combination with a space error model of the machine tool, an improved Sobol global sensitivity analysis method is provided, a first-order sensitivity coefficient and a global sensitivity coefficient are solved, and key geometric errors of the numerical control machine tool are analyzed. By establishing the space error model of the numerical control machine tool and the improved Sobol global error sensitivity analysis model, a new design and improvement concept of the numerical control machine tool is provided, and the problem of machining precision of the numerical control machine tool is solved.

The invention provides a multi-legged robot collaborative mapping method and system based on RTK, and relates to multi-robot collaborative mapping, and the method comprises the steps: constructing a hybrid processing architecture; various sensor data of the foot type robot are obtained, and timestamp alignment is carried out on the various sensor data; calculating a rotation matrix in an initial pose homogeneous transformation matrix between the foot robots and X and Y axis values in a translation vector; the height of a robot body is obtained, and a complete homogeneous transformation matrix between the foot type robots is determined; laser SLAM real-time mapping is executed through the distributed sensing layer; transmitting the local point cloud map of each foot type robot to a data center processing host; performing coarse splicing on the local point cloud map; in the centralized decision-making layer, error compensation is carried out on drifting of the foot-type robot, then fine correction is carried out on the spliced point cloud map, and a global consistency point cloud map is output. According to the method, the accuracy of map splicing is effectively improved, and the calculation amount of mapping is reduced.

The invention relates to the technical field of machine tool precision analysis, in particular to a multi-axis machine tool part quality tracing method and system. The method comprises the steps that a time sequence cutting point location data set and corresponding label data in the part machining process are acquired; constructing a bidirectional long-short term memory network model based on an attention mechanism; on the basis of a multi-body system theory, a homogeneous transformation matrix is adopted to represent a pose transformation relation between adjacent topological structures of the machine tool, and a multi-axis machine tool space error model is established; and inputting the motion axis feed amount of the defect cutting point identified by the defect classification model into the geometric error sensitivity analysis model, quantifying the contribution degree of the geometric error to the machining error, and obtaining a part quality traceability result based on a statistical result of the machining process. According to the method, accurate traceability analysis of the multi-axis machine tool from defect cutting points to key geometric errors is realized, weight information of influence of various geometric errors on machining precision is provided, and optimization design of machine tool precision is effectively guided.

The invention discloses a robot space attitude resolving method and system based on a multi-light-spot image. The resolving method comprises the following steps: collecting laser spot image information emitted by a plurality of laser displacement distance measuring sensors through a depth camera; an internal reference matrix K from an image coordinate system to a camera coordinate system and a homogeneous transformation matrix from the camera coordinate system to a mechanical arm tail end tool coordinate system are obtained; solving to obtain extrinsic parameters of multiple groups of laser displacement distance measuring sensors relative to a camera coordinate system; and constructing a rotation matrix from a mechanical arm tail end tool coordinate system to a plane coordinate system by combining the homogeneous transformation matrix, extrinsic parameters of a plurality of groups of laser displacement distance measuring sensors relative to a camera coordinate system and laser distance measuring distance values. The space coordinates of the plane where the multiple laser spots are located can be solved through the laser ranging distance values and the external parameters, and the problems that the attitude calculation error is large, the attitude adjustment effect is poor and the like due to the fact that the image recognition quality is poor due to the fact that a traditional pure vision scheme is prone to being affected by light rays, shooting angles and other factors can be solved.

The present application belongs to the technical field of gear machining, and discloses a forming milling machining method of spherical gear. The method uses a disc-shaped milling cutter to machine the spherical gear, and the steps are as follows: firstly, the parameters of the spherical gear are given, the tooth surface equation of the spherical gear is derived, and the tooth surface normal vector is calculated; secondly, the forming milling machining coordinate system of the spherical gear is established, and the homogeneous transformation matrix between the coordinate systems is derived, and then the contact line equation between the spherical gear and the disc-shaped milling cutter is derived; then, the cross-sectional profile of the disc-shaped milling cutter is calculated, and then the rotary surface of the disc-shaped milling cutter is obtained; finally, the spherical gear is formed and milled by using the calculated disc-shaped milling cutter, and after machining an annular gear slot, the disc-shaped milling cutter is rotated by a certain angle to continue machining the next gear slot. The forming milling machining method of the spherical gear proposed by the present application can realize the machining of the spherical gear without relying on special gear machining machine tools, and the disc-shaped milling cutter used has a simple structure and is easy to manufacture.

The invention discloses an IMU (Inertial Measurement Unit) drift error calibration method for long-term attitude measurement of an industrial robot. The IMU drift error calibration method comprises the following steps: deducing a pose matrix between a robot base coordinate system and a navigation coordinate system; deriving a pose matrix between a robot tail end coordinate system and a sensor body coordinate system; kinematics modeling is conducted on the industrial robot, and motion parameters are obtained; deriving a homogeneous transformation matrix between two adjacent joints according to the kinematic parameters; acquiring actual joint corner information when the robot runs and actual information measured by a tail end IMU when the robot runs; wherein the actual information of the IMU comprises actual angular velocity information and actual attitude information; the collected actual joint rotation angle information is substituted into the homogeneous transformation matrix, theoretical attitude information of the tail end of the robot is obtained, and the theoretical attitude information is converted into a navigation coordinate system; converting the theoretical attitude information converted into the navigation coordinate system into an Euler angle; designing a Kalman filter; according to a Kalman filter, data fusion of theoretical attitude information and actual attitude information is completed, and calibration is achieved. According to the method, joint angle information, the industrial robot end theoretical attitude obtained through forward kinematics calculation of the kinematics model and actual attitude data obtained through IMU measurement are fused, and accumulated error calibration formed in long-term attitude measurement of the IMU is achieved.

The invention discloses a time-varying precision customized fine self-healing method based on virtual-real homeomorphic evolution, and relates to the technical field of machine tool precision control, and the method comprises the steps: constructing a homogeneous transformation matrix according to the position relation of all parts of a machine tool, and obtaining an error unified mathematical model based on a small error theory; then, a measuring device is used for obtaining small errors of movement and rotation angles and reverse clearance errors of a translation shaft; and then constructing a virtual-real homeomorphic digital twin platform, and integrating the model and the error to realize virtual-real error homeomorphic evolution. Acquiring real-time sensing data, predicting and evaluating precision time variation in combination with a time sequence, and driving virtual-real synchronous evolution to generate a self-healing path; finally, positioning deviation is determined according to path start and end error changes and an error model, and the path is transmitted when a preset value is exceeded; and the time delay defect of traditional real-time compensation is relieved through advanced prediction / deduction. According to the method, unified error modeling of the multi-axis machine tool is achieved, the defect of reverse gap neglect is overcome, errors caused by compensation lag and blind compensation are solved, and the machining precision is guaranteed.

The invention discloses a can foreign matter detection method and system based on intelligent sensor fusion, and relates to the technical field of visual detection. Data collection is carried out through multiple intelligent sensors to obtain multi-source can data; carrying out spatial registration and data enhancement by adopting a homogeneous transformation matrix; extracting X-ray density features by using 3D ResNet-50, extracting texture features by using improved ViT-Base, extracting spectrum time sequence features by using 1D CNN-LSTM, and extracting point cloud space features by using Point Net + +; distributing fusion features through a cross-modal attention mechanism and a dynamic weight; calculating a global anomaly probability and improving YOLOv7 output local anomaly confidence in combination with an SVDD model; fusing global and local results based on a Bayesian network to generate a comprehensive confidence coefficient, and triggering a PLC elimination mechanism when the comprehensive confidence coefficient exceeds a threshold value; the problems of single detection dimension, limited feature expression ability, poor environmental adaptability and poor detection precision in can foreign matter detection are solved.

The invention discloses an intraoperative ultrasonic registration method and system for direct positioning of a bone operation, belongs to the technical field of surgical navigation, and aims to solve the problem of displacement data deviation caused by relative displacement of skin and bones due to physiological movement. Comprising the following steps: acquiring preoperative three-dimensional image data corresponding to an operative region; acquiring a plurality of pieces of three-dimensional ultrasonic image data corresponding to the surgical area; acquiring position data corresponding to the three-dimensional ultrasonic image data; according to the multiple pieces of three-dimensional ultrasonic image data, the position data corresponding to the three-dimensional ultrasonic image data and the preoperative three-dimensional image data, a homogeneous transformation matrix is obtained; acquiring real-time ultrasonic image data acquired by an ultrasonic probe fixedly attached to the surgical area; acquiring real-time position data of an ultrasonic probe for acquiring the real-time ultrasonic image data; according to the homogeneous transformation matrix and the real-time position data, the real-time ultrasonic image data and the preoperative three-dimensional image data are registered, and in-vivo skeleton structure displacement data are estimated.

The invention relates to a contact type machining robot motion planning method based on deep reinforcement learning, and the method comprises the steps: firstly, building position-level forward and inverse kinematics models between the tail end state vector of a six-degree-of-freedom mechanical arm and the motion vectors of a base and a joint based on DH parameters and a homogeneous transformation matrix; secondly, constraints such as a machining task, a feeding speed and a feeding angle are comprehensively considered, and a multi-stage reward function is designed to plan operation parameters; secondly, designing a network structure based on a double-delay depth deterministic strategy gradient, and performing three-level optimization on the algorithm; and finally, designing a planning process according to a network structure, and applying the planning process to a planning task of an intelligent contact type machining mechanical arm system by taking grinding and polishing machining as an example. According to the method, real-time optimization and control of various machining parameters of the contact type machining process can be realized through autonomous learning, and the method can be used for intelligent safety planning under the dynamic coupling condition of the contact type machining process.

The invention discloses a transient signal detection and sorting method based on time-frequency image deep learning, and relates to the technical field of signal sorting. Based on a hypergraph model and core polymorphic sites, the hypergraph model is divided into a plurality of independent substructures by analyzing the distribution characteristics of hyperedge dynamic weights, independent modal analysis is performed on each substructure, and a transient signal detection and sorting result is obtained. Extracting a time-frequency distortion coefficient matrix and an interference response mode matrix; inputting the multi-modal data into a multi-modal network to generate a homogeneous transformation matrix; and based on the homogeneous transformation matrix, deriving attitude error parameters and translation error parameters, feeding back the attitude error parameters and the translation error parameters to a feature input layer and a decision layer of the multi-modal deep learning detection and sorting model, completing dynamic correction of model parameters, and detecting and sorting real-time signals again after model correction. According to the sorting method, the detection rate and sorting accuracy of transient signals are remarkably improved in real-time processing, and meanwhile, the false alarm and leak detection risks are reduced.

The invention relates to a pose compensation method and system for an end effector of a heavy-load mechanical arm. The pose compensation method comprises the following steps: respectively constructing finite element models for a plurality of connecting rod units forming the mechanical arm; solving each finite element model to obtain deformation of the load end under different stress and rod length, and constructing a mapping model of the corresponding deformation of each connecting rod unit under different stress according to the deformation; calculating the actual stress of each connecting rod unit based on the actual load of the actuator and the structural parameters of the connecting rod; calculating the deformation of the connecting rod corresponding to each connecting rod unit through a mapping model of the deformation and the actual stress; according to the forward kinematics homogeneous transformation matrix, the offset of the end effector caused by the deformation of each connecting rod when each connecting rod deforms independently is obtained; and the offsets of all the actuators are accumulated, the compensation amount used for pose compensation of the end effector is generated, pose compensation of the end effector of the mechanical arm is achieved, and the working efficiency and accuracy of the mechanical arm are improved.

The application relates to a picking robot orchard operation area division method, device, equipment and medium, which comprises the following steps: determining the D-H standard parameters corresponding to each joint in a mechanical arm, obtaining the homogeneous transformation matrix of each joint in the mechanical arm according to the D-H standard parameters, determining the position and posture of the end effector of the mechanical arm, and determining the picking area of the mechanical arm according to the position of the end effector by using a probability density function; determining the reachable area corresponding to each grape cluster based on the grape cluster coordinates corresponding to each grape cluster and the picking area of the mechanical arm, and dividing the operation area of the picking robot in the target grape orchard according to the overall distribution information of each grape cluster and the reachable area corresponding to each grape cluster, so as to complete the picking robot orchard operation area division of the picking robot. The application can pay attention to the gathering area with grape distribution for picking, and greatly improves the precision of automatic grape picking.

The invention discloses a robot non-invasive tracing motion compensation method and system based on vision, and the method is realized through a robot non-invasive follow-up control system, and comprises the following steps: carrying out the calibration through hands and eyes; a homogeneous transformation matrix from an NDI coordinate system to a mechanical arm base coordinate system and a homogeneous transformation matrix from a camera coordinate system to a camera Marker coordinate system are obtained; the RGB-D camera is used for detecting the coordinates of the center point of the RoI frame in the operation area, and the coordinates of the center point of the RoI frame in the mechanical arm base coordinate system are obtained through homogeneous transformation matrix conversion; and a position servo control instruction is sent to the mechanical arm, and the mechanical arm is driven to conduct motion compensation on the tail end according to the operation area position change detected by the RGB-D camera. According to the motion compensation method, non-invasive tracking can be achieved, the operation space is saved, the wound surface can be reduced, wounds of the second operation area are avoided, and the problem that tracking fails due to the fact that the Marker in the operation area is prone to being shielded can be solved.

The application discloses a kind of wind turbine single blade installation process dynamic response analysis methods considering wind load action, comprising: according to the wind turbine blade installed and construction sea area, establish blade airfoil database and turbulent wind field database;When the position of lifting point and the position of tappet vertex are described, homogeneous transformation matrix is used to describe the motion of each mechanism of ship-mounted crane;When the position of lifting point is described, the position of tappet vertex is described, and the position of blade centroid in space is described by lifting point position, sling length and generalized coordinate;Wind load at each time point in the process of blade installation is obtained;The sling tension in the process of blade installation is obtained;The dynamics model of single blade installation system is established by using Lagrange method;The method can accurately analyze the coupling motion response between crane ship, ship-mounted crane, sling and blade in the installation process, and simultaneously by selecting generalized coordinate and modeling and analyzing the system based on energy method, the system modeling efficiency is effectively improved.

The invention is suitable for the technical field of magnetic resonance, and provides a robot registration method and device, control equipment and a storage medium, and the method comprises the steps: obtaining a target magnetic resonance image of a development marker installed on a to-be-registered robot; performing image segmentation processing based on an adaptive threshold algorithm on the target magnetic resonance image to obtain target point cloud data of the development marker; according to the target point cloud data, determining position information of the developing marker in a magnetic resonance coordinate system corresponding to the magnetic resonance equipment; according to the position information, determining a target homogeneous transformation matrix of an end tool coordinate system of the robot under a magnetic resonance coordinate system; and the registration operation of the robot is completed by using the target homogeneous transformation matrix. According to the method, robot registration in a magnetic resonance environment is realized, noise reduction processing of the magnetic resonance image is realized in combination with an adaptive threshold algorithm, the determination accuracy of the point cloud data of the developing marker is improved, and the registration accuracy of the robot is further improved.

The application provides a kind of multi-legged robot cooperative mapping method and system based on RTK, it is related to multi-robot cooperative mapping, method includes: constructing mixed processing architecture;Obtain the sensor data of each foot type robot, and each sensor data is timestamped alignment;Calculate the rotation matrix in the initial pose homogeneous transformation matrix between foot type robots and the X, Y axis value in the translation vector;Get body height, determine the complete homogeneous transformation matrix between foot type robots;Through distributed perception layer, laser SLAM real-time mapping is executed;The local point cloud map of each foot type robot is transmitted to data center processing host;Local point cloud map is roughly spliced;In centralized decision-making layer, the drift of foot type robot is compensated for error, and then the spliced point cloud map is corrected, and the global consistent point cloud map is output.The application effectively improves the accuracy of spliced map, and reduces the calculation amount of mapping.

The invention relates to a symmetric calibration method and device for symmetric workpiece point locations, and the method comprises the steps: obtaining first point cloud data of a first workpiece under a mechanical arm base coordinate system, and carrying out the symmetric transformation processing of the first point cloud data to generate virtual second point cloud data of a second workpiece; the mechanical arm six-dimensional point location of the first workpiece is converted into a first homogeneous transformation matrix, a second homogeneous transformation matrix is obtained based on the first homogeneous transformation matrix and a hand-eye matrix of a two-dimensional camera on the mechanical arm, and then an initial mechanical arm six-dimensional point location of a second workpiece is obtained; symmetrically processing the first point cloud data in combination with the internal reference of the camera, converting the first point cloud data into virtual third point cloud data of the second workpiece, and adjusting the second homogeneous transformation matrix based on the deviation between the second point cloud data and the third point cloud data to obtain a third homogeneous transformation matrix; and real point cloud data of the second workpiece are obtained through the three-dimensional camera, and the real mechanical arm six-dimensional point location of the second workpiece is determined in combination with the third point cloud data and the third homogeneous transformation matrix.

The invention discloses a five-axis numerical control machine tool machining error on-line compensation system based on deep learning, relates to the technical field of machine tool error compensation, and solves the problems that position correction difference is caused by neglecting a complex coupling relation between rotating shafts and linear shafts of different types of machine tools, and the error compensation accuracy is high. And deep learning models for machine tools with different requirement types are not designed, so that the model prediction precision is influenced. Determining the coupling relation of each axis error through the HTM homogeneous transformation matrix of the machine tool kinematic chain; aiming at kinematic chain structure differences of different types of machine tools, a differential correction strategy is designed. And constructing a zero drift time sequence model in combination with a historical drift trend term and a random error term predicted by the LSTM. Geometric error terms strongly related to machining errors are screened through grey correlation analysis, then the weight of the geometric error terms is quantified through random forest, the accuracy of key error recognition through a combinatorial algorithm is improved, and the model prediction precision is improved by training fusion models for machine tools with different requirement types.

A kinematics solving method of a cantilever type crushing device, comprising: establishing a link coordinate system of a three-link robot and a homogeneous transformation matrix of an end effector according to obtained DH parameters, and obtaining robot conversion equation sets according to link lengths of a boom, a stick and the end effector and motion pairs of each part; reading current driving space coordinates from displacement sensors of hydraulic cylinders; calculating current pose space coordinates by using the robot conversion equation sets, the link constraint equation sets and the current driving space coordinates; obtaining target pose space coordinates generated after the crushing device is moved to a target working position on the basis of the current pose space coordinates; and solving target driving space coordinates by using the robot conversion equation sets, the link constraint equation sets, the current and target pose space coordinates. The scheme can improve the control precision and motion rate of the cantilever type crushing device, and make the cantilever type crushing device motion more flexible.

This invention discloses a comprehensive error decoupling compensation method for a five-axis CNC polishing machine, comprising: 1. establishing homogeneous transformation matrices for the workpiece kinematic chain and tool kinematic chain of the five-axis polishing machine under ideal and error conditions; 2. establishing a comprehensive error transformation matrix between the coordinate systems of each kinematic pair of the five-axis polishing machine based on the homogeneous transformation matrix; 3. spatially decoupling the comprehensive error transformation matrix of the five-axis polishing machine based on the small error compensation motion assumption, and establishing a comprehensive error compensation model for the five-axis polishing machine to achieve comprehensive error compensation. This invention, through kinematic analysis of a five-axis polishing machine, uses the homogeneous coordinate transformation method to establish a comprehensive error transformation matrix between the coordinate systems of each kinematic pair of the machine tool, spatially decoupling the position and direction error compensation of each motion axis of the five-axis polishing machine, and establishing a comprehensive error compensation model for the five-axis polishing machine, providing a theoretical basis for real-time comprehensive error compensation of a five-axis CNC polishing machine.

The invention discloses a heavy-load high-precision mobile positioning robot, which is characterized in that a plane adjusting module is a 3-PRR parallel mechanism, has three degrees of freedom of X-direction translation, Y-direction translation and rotation around Z, and provides unconstrained vertical support and transverse guidance through a plurality of groups of bull eye bearings, a vision measurement module adopts a monocular vision measurement technology, and the measurement precision is high. And a hand-eye calibration matrix is combined to solve a pose error, accurate coupling with the 3-PRR parallel mechanism is realized, and pose deviation information is fed back to the 3-PRR parallel mechanism to complete accurate pose compensation. According to the method, a homogeneous transformation matrix of a camera coordinate system and a 3-PRR parallel mechanism static platform coordinate system is introduced, a quantized pose transformation relation is established, accurate calculation and compensation of AGV pose deviation are achieved, meanwhile, the standardized detection step of positioning precision is added, the compensation precision is quantitatively detected through a laser tracker, and the accuracy of AGV pose deviation is improved. And a theoretical basis and a quantitative index are provided for positioning precision verification.