A new robot collision detection method: A modified nonlinear disturbance observer based-on neural networks

Xu, Tian; Fan, Jizhuang; Fang, Qianqian; Zhu, Yanhe; Zhao, Jie

doi:10.3233/jifs-179392

Cited by 13 publications

(10 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The algorithm is constructed using the vector analysis based on the model of the minimum distance between the truncated cone generatrix and the cylinder axis. Compared with the method given in the literature, [67][68][69][70][71][72][73][74][75][76][77][78] the cone-cylinder collision detection algorithm proposed in this study is a novel method used in orthopedics to detect collision between the external fixator and the limb during the correction. Simplifying the tibial limb into a mathematical model, and expressing the closeness between the struts and limb in a mathematical perspective will promote the development of many research fields including robotics and orthopedics.…”

Section: Discussionmentioning

confidence: 99%

Influence of parameter deviation on the closeness of the tibial limb and external fixator based on a novel collision detection algorithm

Zuo

et al. 2021

Numer Methods Biomed Eng

View full text Add to dashboard Cite

The Ortho‐SUV frame (OSF) is a hexapod external fixator widely applied in orthopedics deformity correction. The possibility of collision between OSF's struts and the soft tissue is an essential but overlooked issue. To avoid the issue, a novel collision detection algorithm is established based on a cone‐cylinder model of the tibial limb‐strut interaction for detecting the closeness of the tibial limb and external fixator. The algorithm is constructed using the vector analysis based on the model of the minimum distance between the truncated cone generatrix and the cylinder axis. The motion simulation is performed on the overall alignment through the Solidworks‐motion module to verify the feasibility of the algorithm. Subsequently, the installation parameter deviations of the bone‐fixator system are described to investigate the influence of orientation and position deviation on the closeness of the tibial limb and external fixator through the numerical method. The investigation results show that the orientation deviation γ (around the z‐axis), the position deviation τ1 and τ2 (along the x and y‐axes, respectively) have greater sensitivity to closeness and the influence of multiple deviations on the closeness has the property of superposition. The proposed algorithm can assist clinicians to strictly design and appraise frame configurations prior to their application to avoid the collision between the external fixator and the limbs during the correction. It has great application significance in the development of computer‐aided correction software.

show abstract

Section: Discussionmentioning

confidence: 99%

Influence of parameter deviation on the closeness of the tibial limb and external fixator based on a novel collision detection algorithm

Zuo

et al. 2021

Numer Methods Biomed Eng

View full text Add to dashboard Cite

show abstract

“…An example of this kind consists of trained neural networks from manipulator's coupled dynamics in human-robot collision detection [79]. A similar approach proposes a modified nonlinear disturbance observer based on neural networks for improving performance compared to robot dynamics models approaches [80]. Ultimately, there are solutions based on the measured currents of each joint.…”

Section: ) Contact Managementmentioning

confidence: 99%

Human-Robot Interaction Review: Challenges and Solutions for Modern Industrial Environments

et al. 2021

View full text Add to dashboard Cite

The demand for collaborative robots is growing in industrial environments due to their versatility and low prices. Thus, more collaborative solutions are emerging for industrial scenarios. However, implementing scenarios where robots work autonomously while synchronizing their operations in a safe industrial environment with shop-floor workers is not easy. To fill the gap existing in the safe implementation of industrial collaborative scenarios, this manuscript presents a review based on five identified challenges that gathers the primary vital aspects to bear in mind while developing applications for them. Thus, a four-level classification is proposed, which collects the identified challenges and the previous developments in the field of human-robot interaction. The five identified challenges pretends to be the missing enabling key for implementing industrial collaborative scenarios in modern industrial plants. Lastly, a discussion and conclusion are exposed to analyze the degree of development in the field and its potential growth.

show abstract

“…There are also generalized momentum observer (GMO) methods that analyze the robot’s momentum change for collision detection [ 30 , 31 ]. Another method based on the momentum observer can be found in [ 32 , 33 ].…”

Section: Problem Formulationmentioning

confidence: 99%

“…However, this is a model-based approach that is highly dependent on the quality of the dynamic robot model. For collision detection for robots with 6 DoF [ 26 , 30 ], the nonlinear noise observer (NDO) method can also be used.…”

Section: Problem Formulationmentioning

confidence: 99%

A Simple Neural Network for Collision Detection of Collaborative Robots

Czubenko

Kowalczuk

2021

Sensors

View full text Add to dashboard Cite

Due to the epidemic threat, more and more companies decide to automate their production lines. Given the lack of adequate security or space, in most cases, such companies cannot use classic production robots. The solution to this problem is the use of collaborative robots (cobots). However, the required equipment (force sensors) or alternative methods of detecting a threat to humans are usually quite expensive. The article presents the practical aspect of collision detection with the use of a simple neural architecture. A virtual force and torque sensor, implemented as a neural network, may be useful in a team of collaborative robots. Four different approaches are compared in this article: auto-regressive (AR), recurrent neural network (RNN), convolutional long short-term memory (CNN-LSTM) and mixed convolutional LSTM network (MC-LSTM). These architectures are analyzed at different levels of input regression (motor current, position, speed, control velocity). This sensor was tested on the original CURA6 robot prototype (Cooperative Universal Robotic Assistant 6) by Intema. The test results indicate that the MC-LSTM architecture is the most effective with the regression level set at 12 samples (at 24 Hz). The mean absolute prediction error obtained by the MC-LSTM architecture was approximately 22 Nm. The conducted external test (72 different signals with collisions) shows that the presented architecture can be used as a collision detector. The MC-LSTM collision detection f1 score with the optimal threshold was 0.85. A well-developed virtual sensor based on such a network can be used to detect various types of collisions of cobot or other mobile or stationary systems operating on the basis of human-machine interaction.

show abstract

A new robot collision detection method: A modified nonlinear disturbance observer based-on neural networks

Cited by 13 publications

References 20 publications

Influence of parameter deviation on the closeness of the tibial limb and external fixator based on a novel collision detection algorithm

Influence of parameter deviation on the closeness of the tibial limb and external fixator based on a novel collision detection algorithm

Human-Robot Interaction Review: Challenges and Solutions for Modern Industrial Environments

A Simple Neural Network for Collision Detection of Collaborative Robots

Contact Info

Product

Resources

About