Calibration of Multi-Robot Cooperative Systems Using Deep Neural Networks

Maghami, Ali; Imbert, Alaïs; Côté, Gabriel; Monsarrat, Bruno; Birglen, Lionel; Khoshdarregi, Matt

doi:10.1007/s10846-023-01867-6

Cited by 13 publications

(4 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wang et al [17] used the product of exponentials (POE) method as the foundation and further compensated for non-geometric errors through a multi-layer perceptron neural network (MLPNN) optimized by the beetle swarm optimization algorithm, achieving efficient calibration of the SIASUN SR210D robot manipulator. Maghami et al [18] proposed a two-step calibration method based on artificial neural networks (ANNs) for a master-slave cooperative robot system. By training two ANN models to compensate for master-slave relative errors and master robot errors, the absolute accuracy of the master robot and relative tracking precision are improved.…”

Section: Introductionmentioning

confidence: 99%

Kinematic Calibration for the 3-UPS/S Shipborne Stabilized Platform Based on Transfer Learning

Xu,

Tian,

Zhang

2024

JMSE

View full text Add to dashboard Cite

The three-degrees-of-freedom (3-DOF) parallel robot is commonly employed as a shipborne stabilized platform for real-time compensation of ship disturbances. Pose accuracy is one of its most critical performance indicators. Currently, neural networks have been applied to the kinematic calibration of stabilized platforms to compensate for pose errors and enhance motion accuracy. However, collecting a large amount of measured configuration data for robots entails high costs and time, which restricts the widespread use of neural networks. In this study, a “transfer network” is established by combining fine-tuning with a Back Propagation (BP) neural network. This network takes the motion transmission characteristics inherent in the ideal kinematic model as prior knowledge and transfers them to a network trained based on the actual poses. Compared with the conventional BP neural network trained by actual poses alone, the transfer network shows significant performance advantages, effectively solving the problems of low prediction accuracy and weak generalization ability in the case of small-sample measured data. Considering this, the impact pattern of the sample number of the actual pose on the effectiveness of transfer learning is revealed through the construction of multiple transfer network models under varying sample numbers of the actual pose, providing valuable marine engineering guidance. Finally, simulated sea-service experiments were conducted on the 3-UPS/S shipborne stabilized platform to validate the correctness and superiority of the proposed method.

show abstract

Section: Introductionmentioning

confidence: 99%

Kinematic Calibration for the 3-UPS/S Shipborne Stabilized Platform Based on Transfer Learning

Xu,

Tian,

Zhang

2024

JMSE

View full text Add to dashboard Cite

show abstract

“…In cases of multi-robot coordination, the relative pose error manifests as the superposition of pose errors of the mono-robots, resulting in even worse relative accuracy compared to their individual accuracy. In load-sharing tasks, relative errors lead to internal force or even damage of the handled parts (Maghami et al , 2023). And in coordinated fiber placement, relative pose errors can result in defects such as gaps and overlaps between the tows laid by different robots (Cao et al , 2023).…”

Section: Introductionmentioning

confidence: 99%

“…Ma et al (2023) adopted incremental extreme learning machines for error prediction, which entails 500 measurements. Maghami et al (2023) introduced an error prediction and compensation approach for dual-robot systems, which uses two networks for errors of master robot and relative errors between two robots, requiring a total of 1,600 data points. These methods demand a substantial volume of measured data to ensure prediction precision.…”

Section: Introductionmentioning

confidence: 99%

A method for predicting relative position errors in dual-robot systems via knowledge transfer from geometric and nongeometric calibration

Cao,

Wang,

Guo

et al. 2024

View full text Add to dashboard Cite

Purpose In a dual-robot system, the relative position error is a superposition of errors from each mono-robot, resulting in deteriorated coordination accuracy. This study aims to enhance relative accuracy of the dual-robot system through direct compensation of relative errors. To achieve this, a novel calibration-driven transfer learning method is proposed for relative error prediction in dual-robot systems. Design/methodology/approach A novel local product of exponential (POE) model with minimal parameters is proposed for error modeling. And a two-step method is presented to identify both geometric and nongeometric parameters for the mono-robots. Using the identified parameters, two calibrated models are established and combined as one dual-robot model, generating error data between the nominal and calibrated models’ outputs. Subsequently, the calibration-driven transfer, involving pretraining a neural network with sufficient generated error data and fine-tuning with a small measured data set, is introduced, enabling knowledge transfer and thereby obtaining a high-precision relative error predictor. Findings Experimental validation is conducted, and the results demonstrate that the proposed method has reduced the maximum and average relative errors by 45.1% and 30.6% compared with the calibrated model, yielding the values of 0.594 mm and 0.255 mm, respectively. Originality/value First, the proposed calibration-driven transfer method innovatively adopts the calibrated model as a data generator to address the issue of real data scarcity. It achieves high-accuracy relative error prediction with only a small measured data set, significantly enhancing error compensation efficiency. Second, the proposed local POE model achieves model minimality without the need for complex redundant parameter partitioning operations, ensuring stability and robustness in parameter identification.

show abstract

“…To reduce the difficulty of kinematic modeling, many researchers have considered using neural networks to depict the relationship between the end effector error and the joint angles of industrial robots [16,17]. Maghami et al proposed a two-step calibration method for a master-slave collaborative robot system based on artificial neural networks (ANNs), using joint angles and output pose errors as the training data [18]. Ma et al employed an incremental extreme learning machine (IELM) to predict the positioning error of an industrial robot and improve its positioning accuracy [19].…”

Section: Introductionmentioning

confidence: 99%

Error Modeling and Parameter Calibration Method for Industrial Robots Based on 6-DOF Position and Orientation

Lao,

Quan,

Wang

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

The positional accuracy and orientation accuracy of industrial robots are crucial technical indicators for determining their applicability in industrial scenarios. However, the majority of current calibration methods for industrial robots only consider positional errors, neglecting the significance of orientation accuracy. This paper presents a more accurate error model and parameter calibration method for industrial robots based on six degrees-of-freedom position and orientation to identify the actual structural parameters. Firstly, based on the modified Denavit–Hartenberg parameters, the transformation errors of the tool coordinate system and measurement coordinate frame were introduced to establish a geometric parameter error model with positional and orientation accuracy as the optimization objectives. Secondly, to address the drawback of falling into local optima when identifying geometric parameters simultaneously, a geometric parameter cross-identification method based on the Levenberg–Marquardt algorithm is proposed. Lastly, the linear relationship between the parameters was analyzed, and a scheme for not calibrating some geometric parameters under specific conditions was given. Simulation results demonstrated that, under the premise of existing transformation errors, the proposed geometric parameter error model can accurately identify the actual structural parameters of industrial robots. After calibration, the positional error at the robot’s flange end decreased from 1.9536 mm to 0.0122 mm, and the orientation error decreased from 1.46 × 10−2 rad to 1.31 × 10−4 rad. Furthermore, compared to identifying the geometric parameters simultaneously, the proposed cross-identification method has a wider convergence range.

show abstract

Calibration of Multi-Robot Cooperative Systems Using Deep Neural Networks

Cited by 13 publications

References 24 publications

Kinematic Calibration for the 3-UPS/S Shipborne Stabilized Platform Based on Transfer Learning

Kinematic Calibration for the 3-UPS/S Shipborne Stabilized Platform Based on Transfer Learning

A method for predicting relative position errors in dual-robot systems via knowledge transfer from geometric and nongeometric calibration

Error Modeling and Parameter Calibration Method for Industrial Robots Based on 6-DOF Position and Orientation

Contact Info

Product

Resources

About