Model Of Industrial Robot Research Articles

Neural network–based transfer learning to improve stiffness modeling of industrial robots with small experimental data sets

Stiffness modeling is an essential subject for the composition of robot control. Accurate stiffness modeling is helpful for improving the control accuracy of industrial robots, particularly under dynamic load circumstances. The classic virtual joint modeling (VJM) method is challenging in predicting the deformation of the end-effector throughout the full workspace due to the nonlinear deformation of the robot joint and its serial articulated structure. This paper proposes a full-space stiffness modeling method for robots based on the integration of a multi-layer perceptual (MLP) model and VJM. To provide enough training data for the MLP model, VJM is used to build a stiffness model with a small set of experimental data to generate 106,400 training data. A model-based transfer learning approach is proposed to improve the model’s accuracy and generalization regarding the difference between generated training data and actual experimental data. The VJM stiffness model is compared with the MLP stiffness model and the existing CNN-based transfer learning model based on the same experimental data. Considering the deformation prediction in the three directions in Cartesian space, the mean absolute error, standard deviation, and maximum error of the MLP model are decreased by at least 24.90%, 14.20%, and 8.50%, respectively, than the VJM. These prediction results demonstrate that the proposed modeling technique can significantly increase the accuracy of robot stiffness modeling, which is essential for position compensation in precise motion control of robots under dynamic load.

An Efficient Integral Approach for Kinematic Reliability Sensitivity Analysis of Industrial Robots

Abstract Assessment of the reliability and reliability sensitivity of positioning accuracy for industrial robots subject to aleatoric and epistemic uncertainties registers a challenging task. This study proposes a new optimized moment-based method for kinematic reliability analysis and its sensitivity analysis, which incorporates the sparse grid (SPGR) technique and the saddlepoint approximation (SPA) method. To start with, the positioning accuracy reliability and its sensitivity models of industrial robots are established via computational optimization techniques and kinematic criteria. The kinematic accuracy reliability and its sensitivity are then calculated. Specifically, the sparse grid technique is adopted to approach the positioning error statistical moments and moment sensitivities. On this basis, positioning accuracy reliability bounds and reliability sensitivity bounds are obtained by the saddlepoint approximation method and optimization techniques. Finally, two practical examples are implemented to demonstrate the proficiency of the currently proposed method against Monte Carlo simulation (MCS) results. The results show that the currently proposed method exhibits superior computational accuracy and efficiency in kinematic reliability and its sensitivity analyses for industrial robots.

Research on Safety Control Technology for Dual Robots Based on Kinematic Models

The industrial robots play an important role in modern industrial manufacturing processes, and their applications are becoming increasingly widespread, greatly improving manufacturing efficiency and significantly enhancing the digitalization level of the manufacturing industry. Therefore, achieving safe control of multi robot collaborative work is of great significance for the stable and efficient operation of robot systems. This article proposes a dual robot safety control technology based on the kinematic model of industrial robots. Firstly, a mathematical model of the kinematic space of the dual robots is constructed using D-H parameters; Then, use recursive traversal algorithm to calculate the minimum distance between each joint arm of the two robots in real time; Finally, the minimum distance between each joint arm will be calculated and compared with the set safety threshold to achieve real-time assessment and control of collision risk during the movement of the dual robots. A dual robot safety control testing platform was built using two KUKA industrial robots, integrated control software was developed, and relevant algorithms were validated. Through experiments, it was verified that this method can evaluate the collision risk of dual robots in real time during motion and effectively control the collision risk during the motion of dual robots.

A workload identification method of industrial robot combining dynamic model and convolutional neural network

The motion joint system of industrial robot has obvious nonlinear characteristics, high dimensional running data and limited number of experimental samples. If mature neural network algorithm is used for identifying the workload of robots, it is easy to have over-learning problems, which seriously restricts the generalization ability of load identification method. In this paper, a load identification method combining the dynamic model of industrial robots with the neural network data model is proposed. A UR5 robot is used for carrying out multiple dynamic load testings to verify the effectiveness of the proposed identification method. Firstly, the workload identification by CNN algorithm is given, and the influence of parameters of prediction model and different neutral network are analyzed. Then the classical dynamic model of industrial robots with multi-degrees of freedom is established. The identified workload by dynamic model is also analyzed. At last, the deterministic information such as velocity and displacement is extracted from the calculation results of the dynamic model as the initial anchoring value. Then convolutional neural network is applied for compensating the residual highly nonlinear information. An improved mixing combination method is also proposed. This method can effectively deal with the interaction of different types of information in the data, and preliminarily cooperate the dynamic model and convolutional neural network. This provides a basic framework and method for solving the problem of parameter identification in multi-degree-of-freedom systems with small samples.

Development of Deep Learning for Power Energy Optimization in the Industrial Robot System

This paper established power consumption modeling and motion estimation optimization of industrial robots. We also studied factors affecting the use of electrical energy, such as friction, torque, and electric current. The energy consumption parameters of each coupling can be quantified through the Deep Learning (DL) technique, Scaled Conjugate Gradient (SCG) estimation, or Simulation and experimentation based on the movement posture of a given robot dynamic model to control the robot operation. The robot dynamic model parameters can be identified and expressed in mathematical equations. Electrical energy consumption estimates were analyzed using the SCG technique to compare with the Nonlinear Least Squares (NLS) method using a large dataset of approximately 60,000 samples. The results showed accurate parameter prediction and electrical energy consumption estimation of the robot locomotion pose. The maximum errors in the SCG and NLS methods were 0.89% and 1.54%, respectively. It indicated that the electric energy consumption model using the SCG estimation method is more efficient than the NLS method.

Accelerated Degradation Reliability Modeling Based on Multiple Uncertainty Wiener Processes

Numerous scholars' studies usually adopt the Wiener process to establish industrial robot performance degradation models and randomize the drift coefficients to characterize the differences in individual degradation rates, but they do not take into account the influence of measurement errors on the accuracy of the models. Aiming at the above problems, combining the Wiener process and the accelerated degradation model, the accelerated degradation reliability model of industrial robots based on the multi-uncertainty Wiener process is established, and by adding the factor of measurement error, the model is able to more accurately characterize the influence of external factors on the degradation process of industrial robots. Bayesian theory, Markov chain Monte Carlo method and Bootstrap method are used to estimate the unknown parameters in the model. Finally, the accuracy of the proposed parameter estimation method and model is verified by simulating the industrial robot performance degradation data with Monte Carlo method.

A Neural Network Separation Approach for the Inclusion of Static Friction in Nonlinear Static Models of Industrial Robots

A Neural Network Separation Approach for the Inclusion of Static Friction in Nonlinear Static Models of Industrial Robots

Optimal design for compliance modeling of industrial robots with bayesian inference of stiffnesses

In this paper a cost and time efficient approach to setup a compliance model for industrial robots is presented. The compliance model is distinctly determined by the gear’s stiffness parameters which are tuned by an optimal design of experiments approach. The experimental setup consists of different poses of the robot’s axes together with the applied force at the tool center point (TCP). These robot poses represent together with defined forces the experimental setup where the deviation of the robot under defined force is measured. Based on measurements of the displacement of the TCP the stiffness parameters for the compliance model are estimated and afterwards validated in new experiments. The efficiency of this approach lies in the reduced amount of experiments that are needed to identify the stiffness parameters that are parameters inherent to the compliance and the less complex experimental setup.

Fault diagnosis model of multi-axis industrial robot based on triplet network

Fault diagnosis is an important link in intelligent development of industrial robots. Aiming at the problem of weak fault diagnosis performance caused by insufficient training samples, a fault diagnosis model based on triplet network is proposed. Firstly, we combine the multiscale convolutional neural network (MSCNN) with channel attention networks (squeeze-and-excitation network, SENet), and use it to construct a triple sub-network structure MS-SECNN, which can adaptively extract features from the original fault signal. Then, the feature similarity is calculated by triplet loss in the low dimensional space to realize the fault classification task. The experiments are based on the real industrial robot operation data set. In this model, we use Few-shot learning strategy to test the diagnostic performance under small samples, and compare it with WDCNN, FDCNN and MSCNN models. Experimental results show that the proposed model has more effective fault classification ability under small samples. In addition, when the training sample size is 1400, the average accuracy of MS-SECNN reaches 99.21%.

Analysis of the base motion model of industrial robots moving

Development of mathematical models and algorithms for optimal control of the functioning of industrial robots on a movable base to ensure the accuracy of the trajectory of movement and positioning is considered. An equation for the functioning of industrial robots on a movable base in the implementation of a complex spatial operation obtained and, on its basis, a mathematical model of optimal control is developed.

Estimating pose-dependent FRF in machining robots using multibody dynamics and Gaussian Process Regression

Frequency Response Functions (FRF) of the robot at its TCP are essential for modeling and suppression of industrial robots’ vibrations during machining operations. Because the robot’s FRF change by posture, measuring the FRF experimentally (e.g. by modal testing) is not efficient, and predictive models are needed to obtain the FRF in arbitrary postures. Multibody dynamics models are efficient in estimating pose-dependent FRF, but they usually include a large number of inertial and joint elastic parameters that must be identified experimentally. Moreover, while the inertial parameters in the model are pose-independent, the joint elastic parameters vary significantly by posture. In this work, we present a new parameter identification method that improves the identifiability of multibody models by systematically imposing constraints according to the robot’s rigid-body dynamics as well as its physical and geometrical properties. We then use Gaussian Process Regression (GPR) to model the variation of the joint elastic parameters by posture. This new approach in combining multibody modeling with data-driven modeling (GPR) is more generalizable than purely data-driven methods in predicting pose-dependent FRF because it considers the known physics of the system in predictions. The presented method is used to develop a flexible-joint multibody model for a KUKA KR90 robotic arm with a machining end-effector, and the FRF predicted by the model are compared to the FRF measured by impulse hammer tests to validate their accuracy. • A hybrid approach is presented to model pose-dependent FRF in machining robots. • New optimization problem is formulated to identify vibration model parameters • Variations in the joints’ dynamics are modelled by Gaussian Process Regression elastic parameters by its posture.

Sim2real kinematics modeling of industrial robots based on FPGA-acceleration

It is challenging to apply robotic inverse kinematics directly in simulation to real robot motion due to the multi-configuration problem, robot end deviation, and delay. To this end, this paper proposes a industrial robot modeling method for “Simulation to Real” to obtain the real joint angle vector given real end-effector target pose . In the first stage, the joint angle in simulation corresponding to the real end-effector pose, is calculated by improved Newton–Raphson method. In the second stage, a data-driven ResNet model, which is trained on enough simulated joint angles and end-effector pose obtained in the first stage, is used to acquire the end-effector pose in simulation. The joint angles in real environment can be obtained by the simulated end-effector pose in execution stage. The experiment takes a six-axis industrial robot as a case study. The results verify the feasibility and real-time performance of the inverse kinematics method in robotic motion from simulation to real environment. • A sim2real modeling method for the kinematics of industrial robots is proposed. • An FPGA-accelerated inverse kinematics model based on iteration is built. • An FPGA-accelerated forward kinematics model in simulation based on ResNet is built.

Review of Industrial Robot Stiffness Identification and Modelling

Due to their high flexibility, large workspace, and high repeatability, industrial robots are widely used in roughing and semifinishing fields. However, their low machining accuracy and low stability limit further development of industrial robots in the machining field, with low stiffness being the most significant factor. The stiffness of industrial robots is affected by the joint deformation, transmission mechanism, friction, environment, and coupling of these factors. Moreover, the stiffness of a robot has a nonlinear distribution throughout the workspace, and external forces during processing cause irregular deviations of the robot, thereby affecting the machining accuracy and surface quality of the workpiece. Many scholars have researched identifying the stiffness of industrial robots and have proposed methods for improving the performance of industrial robots, mainly by optimizing the body structure of the robot and compensating for deformation errors with stiffness models. This paper reviews recent research on the stiffness modelling of industrial robots, which can be broadly classified as finite element analysis (FEA), matrix structure analysis (MSA), and virtual joint modelling (VJM) methods. Each method is studied from three aspects: algorithms, implementation, and limitations. In addition, common measurement techniques have been introduced for measuring deformation. Further research directions are also discussed.

High-accuracy prediction and compensation of industrial robot stiffness deformation

Industrial robots (IRs) are promising options for machining large complex structural parts due to the higher flexibility, larger operating space, and lower cost compared with multi-axis machine tools. However, the relatively low posture-dependent stiffness and large stiffness deformation of IRs significantly deteriorate the contour accuracy of milling in which the cutting force is large generally. It is very complex to achieve a precise stiffness model and predict stiffness deformation of IRs because of the joint clearance, drift of zero-position, and other nonlinear factors. The conventional stiffness model of IRs only takes each joint as a constant linear torsion spring into consideration and ignores other difficult-to-model factors, which leads to low-accuracy identified results and thereafter induces deformation prediction errors. The data-driven approach can be used to obtain an accurate stiffness and deformation model, but a large amount of experimental data is required and it will cost enormous time and effort. In order to circumvent the experimental data deficiency and difficult-to-model issue, a simulation-driven transfer learning method named Adaptive Domain Adversarial Neural Network with Dual-Regressions (ADANN-2R) is designed for robot deformation prediction. Amounts of coarse deformation data, which are generated by the conventional stiffness model, are regarded as source data. And few real deformation data, which are obtained by deformation experiments, are regarded as target data. The Dual-Regressions are designed after the feature extractor, and the weighting parameters are adjusted adaptively to tackle the different magnitude of the regression loss and domain discrimination loss. The ADANN-2R aligns the simulated source data and real target data to perform adversarial training, and an accurate target deformation predictor is achieved. Experimental results indicate that the proposed ADANN-2R can obtain high-accuracy prediction with few real data compared with the conventional stiffness model. Compared with the path without deformation compensation and the pre-compensated path using the conventional stiffness model, the maximum position error of the pre-compensated path using the proposed ADANN-2R is reduced by 78.12% and 32.45%, respectively.

Research on Fault Diagnosis Technology of Industrial Robot Operation Based on Deep Belief Network

Fault diagnosis technology is the science of identifying the operating state of a machine or unit, and it studies the response of the change in the operating state of the machine or unit in the diagnostic information. It can give an early warning to the failure state of the machine and stop the machine before a major failure occurs so as to protect the life safety of the on-site staff and avoid huge economic losses to the enterprise. For mechanical equipment, fault diagnosis consists of three main links: fault detection; fault identification; and fault classification. Aiming at the problems that need to be solved in the fault diagnosis of industrial robots, this paper adopts a data-driven intelligent diagnosis method to establish a fault diagnosis model of industrial robots based on Deep Belief Network (DBN) and DSmT theory. Firstly, based on wavelet transform and information energy entropy correlation theory, the vibration signal of industrial robot is extracted, and the energy entropy normalized eigenvector is established. Then, the energy entropy normalized feature vector is divided into training set and test set to complete the creation of DBN network model. Finally, using DSmT theory to carry out decision-making fusion, a fault diagnosis model for industrial robots is established, and experiments are carried out on the K-R-R540 robot to verify the applicability of the established fault diagnosis model. It is proved by experiments that the industrial robot fault diagnosis model based on the deep belief network can meet the requirements of the recognition accuracy of robot faults, and the model will perform poorly when the faults coexist with multiple faults.

Robot tangram assembly line based on vision

With the development of smart manufacturing technology, the cost of industrial robots is reduced and reliability enhanced. Applying several robots for automated production lines will be more popular. The extended application of industrial robots can improve the level of automation in industrial production and break the boundaries of technology and costs. The project focuses on "Tangram assembly line". This project takes "tangram assembly line" as the research object, and uses the kinematics model and path planning method of industrial robots to complete the identification, assembly and disassembly of tangram. "Tangram" simulated industrial products, based on two-dimensional vision to carry out image processing, identify the position and Angle of the object information, through writing robot program, the final completion of the object splicing and disassembly work. CoppeliaSim Edu software simulation results prove that the project can be efficient, high quality to complete the expected task.

Design and modeling industrial intelligent robot flexible hand

The article describes a description of a fundamentally new design, mathematical model and experimental research of a flexible arm with an anthropomorphic gripper for an industrial robot. The advantage of the proposed design of the robot arm in comparison with the known traditional technical solutions is achieved as close as possible to the functions of the human arm. This property significantly increases the versatility of the robot arm when performing various technological operations. Another difference from the known models of industrial robots is the presence of an anthropomorphic gripping device in the flexible arm, which allows you to service products with different shapes and arbitrary coordinates in space. In addition, the article for the first time proposes a method for calculating the parameters of a new hand and experimental studies of its functioning, which will allow engineers in the field of robotics to create similar designs. The economic effect of the proposed design is that the implementation of the movements of the proposed robot arm does not require separate electromechanical drives for each joint of the kinematic chain of the manipulator. This effect significantly reduces the cost of a robot arm while expanding its technological capabilities.

Dynamic Simulation Modeling of Industrial Robot Kinematics in Industry 4.0

This paper studies the kinematic dynamic simulation modeling of industrial robots in the Industry 4.0 environment and guides the kinematic dynamic simulation modeling of industrial robots in the Industry 4.0 environment in the context of the research. To address the problem that each parameter error has different degrees of influence on the end position error, a method is proposed to calculate the influence weight of each parameter error on the end position error based on the MD-H error model. The error model is established based on the MD-H method and the principle of differential transformation, and then the function of uniform variation of six joint angles with time t is constructed to ensure that each linkage geometric parameter is involved in the motion causing error accumulation. Through the analysis of the robot marking process, the inverse solution is optimized for multiple solutions, and a unique engineering solution is obtained. Linear interpolation, parabolic interpolation, polynomial interpolation, and spline curve interpolation are performed on the results after multisolution optimization in the joint angle, and the pros and cons of various interpolation results are analyzed. The trajectory planning and simulation of industrial robots in the Industry 4.0 environment are carried out by using a special toolbox. The advantages and disadvantages of the two planning methods are compared, and the joint space trajectory planning method is selected to study the planning of its third and fifth polynomials. The kinetic characteristics of the robot were simulated and tested by experimental methods, and the reliability of the simulation results of the kinetic characteristics was verified. The kinematic solutions of industrial robots and the results of multisolution optimization are simulated. The methods, theories, and strategies studied in this paper are slightly modified to provide theoretical and practical support for another dynamic simulation modeling of industrial robot kinematics with various geometries.

Intelligent Technologies and Robotic Machines for Garden Crops Cultivation

The existing models of industrial robots cannot perform technological processes of apple harvesting. It is noted that there is a need for developing special actuators, grippers and new control algorithms for harvesting horticulture products. (Research purpose) The research aimed to develop an intelligent control system for horticulture industrial technologies and robotic techniques for yield monitoring and fruit harvesting. (Materials and methods) The research methodology was based on such modern methods as computer modeling and programming. In particular, the following methods were applied: systems analysis, artificial neural networks theory, pattern recognition, digital signal processing. The development of software, hardware and software was carried out in accordance with the requirements of GOST technical standards. The following programming languages were used: (C / C ++)-based OpenCV library, Spyder Python Development Environment, PyTorch and Flask frameworks, and JavaScript. Image marking for training neural networks was carried out via VGG ImageAnnotator and in Labelbox. The design process was based on the finite element method, CAD SolidWorks software environment. (Results and discussion) An intelligent management system for horticulture industrial technologies has been created based the on the «Agrointellect VIM» hardware and software complex. The concept of the system is shown to be implemented via computer and communication technology, robotic machines, the software for collecting, organizing, analyzing and storing data. The gripper proves to fix an apple gently and holds it securely. Depending on the size, the fruit fixation time is 1.5-2.0 seconds, the fruit maximum size is 85 per 80 millimeters , and its maximum weight is 500 grams. (Conclusions) The developed intelligent control system for industrial technologies based on «Agrointellect VIM» hardware and software complex ensures the efficient real-time processing of information necessary for the design of intelligent agricultural technologies using robotic machines and artificial intelligence systems.

A methodology for industrial robot calibration based on measurement sub-regions

This paper proposes a methodology for calibration of industrial robots that uses a concept of measurement sub-regions, allowing low-cost solutions and easy implementation to meet the robot accuracy requirements in industrial applications. The solutions to increasing the accuracy of robots today have high-cost implementation, making calibration throughout the workplace in industry a difficult and unlikely task. Thus, reducing the time spent and the measured workspace volume of the robot end-effector are the main benefits of the implementation of the sub-region concept, ensuring sufficient flexibility in the measurement step of robot calibration procedures. The main contribution of this article is the proposal and discussion of a methodology to calibrate robots using several small measurement sub-regions and gathering the measurement data in a way equivalent to the measurements made in large volume regions, making feasible the use of high-precision measurement systems but limited to small volumes, such as vision-based measurement systems. The robot calibration procedures were simulated according to the literature, such that results from simulation are free from errors due to experimental setups as to isolate the benefits of the measurement proposal methodology. In addition, a method to validate the analytical off-line kinematic model of industrial robots is proposed using the nominal model of the robot supplier incorporated into its controller.