Robot inverse kinematics via neural and neurofuzzy networks: architectural and computational aspects for improved performance

Raptis, Spyros; Tzafestas, Elpida

doi:10.1080/02522667.2007.10699780

Cited by 6 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The mean error presented in [1] is 0.9 mm, with standard deviation of 2.9 mm implying that are points in the workspace that the error is higher than the one presented in this paper. Finally, the RMS error presented in [8] was less than 2 % which is much greater than the one presented in this paper.…”

Section: Simulated Experiments and Resultscontrasting

confidence: 62%

See 1 more Smart Citation

A Method for the Approximation of the Multiple IK Solutions of Regular Manipulators Based on the Uniqueness Domains and Using MLP

Moulianitis

Kokkinopoulos

Aspragathos

2015

Robotics and Mechatronics

View full text Add to dashboard Cite

In this paper, a method for dividing the training data as well as training MLP systems to obtain the multiple solutions of IKP of regular manipulators is presented. The sets of training data for each system are strictly defined using the concept of uniqueness domains. The training data are obtained by the forward kinematics and the sign of the determinant of the manipulator Jacobian is used for the determination of the uniqueness domains. An illustrative example with a 3 dof robot with known IK solutions is presented for the verification of the proposed approach.

show abstract

Section: Simulated Experiments and Resultscontrasting

confidence: 62%

“…An algorithm [8] which used radial basis function networks and two fuzzy logic methods to determine all the solutions of the IKP was presented. The training data were distinguished manually based on prior knowledge of the multiple configurations corresponding to a given pose of the robot.…”

mentioning

confidence: 99%

A Method for the Approximation of the Multiple IK Solutions of Regular Manipulators Based on the Uniqueness Domains and Using MLP

Moulianitis

Kokkinopoulos

Aspragathos

2015

Robotics and Mechatronics

View full text Add to dashboard Cite

show abstract

“…Feedforward neural networks used in their study had a single hidden layer [13]. Raptis & Tzafestas obtained inverse kinematics of the PUMA 3R manipulator via Neurofuzzy and Neural Networks [23]. Almusawi et.al., proposed a multilayer neural network with 6 input variables to solve inverse kinematics of Denso VP6242 robotic arm [3].…”

Section: Introductionmentioning

confidence: 99%

Visual Servoing Application for Inverse Kinematics of Robotic Arm Using Artificial Neural Networks

Ak¹,

Topuz²,

Erşan³

2019

STUD INFORM CONTROL

View full text Add to dashboard Cite

This paper presents novel approach for a visual servoing application of six axis robotic arm. Basic imageprocessing techniques were used for object recognition and position determination of robotic arm. The inverse kinematics solution of the robot arm was performed with artificial neural networks. Afterwards the robot's inverse kinematics solution was completed, the determined joint-angle values were used to control the robot arm. Performance of radial basis function network (RBF) and multilayer perceptron (MLP) were also compared.

show abstract

“…Two important factors motivated the investigation of better prediction models as the dimensionality of the ANN search spaces was set higher: i.e. the inclusion of the orientation of the robot in the input vector and use of robots with higher DoF [19,[30][31][32][33]. In a recent study, El-Sherbiny et al [19] applied three soft-computing methods -MLP, AN-FIS, and GA -to solve the inverse kinematics of a 5 DoF robotic arm.…”

Section: Survey On Data-driven Methodsmentioning

confidence: 99%

“…But as in the other studies, the results were restricted to a particular robot in a limited workspace. Raptis and Tzafestas [32] [38] also used a Self-Organizing Map (SOM) with different sizes (18x18, 36x36 and 72x72) to learn the inverse kinematics of an Aura robotic arm with only 2 DoF and confined to two well-defined trajectories: linear and planar.…”

Section: Survey On Data-driven Methodsmentioning

confidence: 99%

Use of Jacobians for inverse kinematics of articulated robots : a study on approximate solutions

Demby's¹

View full text Add to dashboard Cite

In the context of articulated robotic manipulators, the Forward Kinematics (FK) is a highly non-linear function that maps joint configurations of the robot to poses of its endeffector. Furthermore, while in the most useful cases these functions are neither injective (one-to-one) nor surjective (onto), depending on the robot configuration -- i.e. the sequence of prismatic versus revolute joints, and the number of Degrees of Freedom (DoF) -- the associated Inverse Kinematics (IK) problem may be practically or even theoretically impossible to be solved analytically. Therefore, in the past decades, several approximate methods have been developed for many instances of IK problems. The approximate methods can be divided into two distinct categories: data-driven and numerical approaches. In the first case, data-driven approaches have been successfully used for small workspace domains (e.g., task-driven applications), but not fully explored for large ones, i.e. in task-independent applications where a more general IK is required. Similarly, and despite many successful implementations over the years, numerical solutions may fail if an improper matrix inverse is employed (e.g., Moore-Penrose generalized inverse). In this research, we propose a systematic, robust and accurate numerical solution for the IK problem using the Unit-Consistent (UC) and the Mixed (MX) Inverse methods to invert the Jacobians derived from the Denavit-Hartenberg (D-H) representation of the FK for any robot. As we demonstrate, this approach is robust to whether the system is underdetermined (less than 6 DoF) or overdetermined (more than 6 DoF). We compare the proposed numerical solution to data driven solutions using different robots -- with DoF varying from 3 to 7. We conclude that numerical solutions are easier to implement, faster, and more accurate than most data-driven approaches in the literature, specially for large workspaces as in task-independent applications. We particularly compared the proposed numerical approach against two data-driven approaches: Multi-Layer Perceptron (MLP) and Adaptive Neuro-Fuzzy Inference System (ANFIS), while exploring various architectures of these Neural Networks (NN): i.e. number of inputs, number of outputs, depth, and number of nodes in the hidden layers.

show abstract

Robot inverse kinematics via neural and neurofuzzy networks: architectural and computational aspects for improved performance

Cited by 6 publications

References 12 publications

A Method for the Approximation of the Multiple IK Solutions of Regular Manipulators Based on the Uniqueness Domains and Using MLP

A Method for the Approximation of the Multiple IK Solutions of Regular Manipulators Based on the Uniqueness Domains and Using MLP

Visual Servoing Application for Inverse Kinematics of Robotic Arm Using Artificial Neural Networks

Use of Jacobians for inverse kinematics of articulated robots : a study on approximate solutions

Contact Info

Product

Resources

About