Kinematic Estimation with Neural Networks for Robotic Manipulators

Theofanidis, Michail; Sayed, Saif; Cloud, Joe; Brady, James; Makedon, Fillia

doi:10.1007/978-3-030-01424-7_77

Cited by 11 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The complex structures of robot arms restrict the implementation of the traditional approach techniques of inverse kinematics solutions [12,13] (algebraic approach, geometric approach, and iterative approach) because it takes a long time and requires complex computations. Many researchers have investigated a solution to the inverse kinematics problem of the n-DOF robot arm using the Adaptive Neuro-Fuzzy Inference System (ANFIS) technique [14,15] and the artificial neural network (ANN) technique [16,17] as an alternative to traditional techniques in addition to the Fuzzy Logic Controller (FLC), which is utilized for the dynamic and kinematic analysis of mobile robot manipulators owing to its adaptive feature [18][19][20]. For the kinematics solution, FLC was applied for the determination of the joints variables as the controller outputs through fuzzy rules with end-effector positions as the controller inputs as explained by Crenganiș et al [21].…”

Section: Introductionmentioning

confidence: 99%

Experimental validation of a neuro-fuzzy controlled mobile manipulator for the inspection of automobile exhaust system

Gaber,

Eldrainy,

Awad

et al. 2024

Preprint

View full text Add to dashboard Cite

This study establishes a new strategy to locate the coordinates of a crack in an automobile exhaust system using a 6 Degree of Freedom (DOF) mobile robot manipulator, utilizing a fuzzy control model and an Artificial Neural Network (ANN). The fuzzy position control model was applied as an alternative to traditional inverse kinematics approaches, and an ANN was used to correct the position of the robot based on a manipulator workspace. A Carbon Dioxide (CO2) sensor was mounted on the base of the mobile robot to detect exhaust gas leakage. Several uncertain manipulator parameters affect precision and performance. These parameters are created from the tolerances and variations in the actuators and controls for production and assembly. To ensure manipulator end-effector location, a fuzzy control model with an Internet Protocol (IP) camera-based observational error model is applied to compensate for the robot parameter errors and improve the location accuracy of the robot manipulator. To demonstrate the effectiveness of the proposed control model, a simulation was implemented during the uncertainty in the robot link length. 6 DOF mobile robot manipulator was equipped at a low cost to help in the experimental validation to establish the robustness of the proposed controller in real-world applications.

show abstract

Section: Introductionmentioning

confidence: 99%

Experimental validation of a neuro-fuzzy controlled mobile manipulator for the inspection of automobile exhaust system

Gaber,

Eldrainy,

Awad

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Applications of Artificial Neural Network for forward kinematics estimation problems are prominent in the literature. Several such approaches are established for various robotic setups such as HEXA Parallel Robot (Dehghani et al, 2008), parallel manipulators (Parikh et al, 2009), 3D cable robot (Ghasemi et al, 2010), 7-DOF Sawyer Robotic Arm (Theofanidis et al, 2018), Delta parallel robot (Liu et al, 2019) model, and so on. Further, improvement was observed with integration of PSO to NN for a 6-DOF parallel robot (Li et al, 2007) to fine-tune backpropagation learning.…”

Section: Introductionmentioning

confidence: 99%

Transfer Learning-Based Artificial Neural Network for Forward Kinematic Estimation of 6-DOF Robot

Kesaba

Choudhury

2022

International Journal of Applied Metaheuristic Computing

View full text Add to dashboard Cite

Transfer Learning (TL) can significantly lower training time and reduce dependency on a large number of target domain datasets. Such an approach is still not exploited for robotic prediction tasks. Currently, a TL based Artificial Neural Network (ANN) is explored and validated for robotic forward kinematics estimation of a 6-DOF robot. The robotic positions are estimated from the available joint angle information. The 6-R MTAB Aristo-XT robot is selected as a case study to generate the target experimental training and testing data for validation of ML techniques. While, the PUMA 560 6-DOF robot is used as a source for prior training of the ANN model. Standard performance measures such as learning error, deviation error and Mean Square Error (MSE) are evaluated and graphical illustrations are presented for fair comparison of the results. Experimental results reveal that, instead of ANN, the TL-ANN is strongly suggested to improve the training time of ANN regressor, and it also reduces the randomness and improves the accuracy as compared to its counterpart.

show abstract

“…To move the position of an end-effector upon a certain trajectory, a combination of angular/linear motions by the motors at each joint provide that path. The equations that connect the position of the end-effector and the angular positions of the joints are called the kinematic equations of the manipulator [2].…”

Section: Introductionmentioning

confidence: 99%

“…A P is the translation of frame A to frame B and A B R is the rotation of frame A to frame B. Using the simple laws of trigonometry, the rotation of each axis is depicted with Equation (2).…”

Section: Introductionmentioning

confidence: 99%

Forward Kinematic Modelling with Radial Basis Function Neural Network Tuned with a Novel Meta-Heuristic Algorithm for Robotic Manipulators

2022

View full text Add to dashboard Cite

The complexity of forward kinematic modelling increases with the increase in the degrees of freedom for a manipulator. To reduce the computational weight and time lag for desired output transformation, this paper proposes a forward kinematic model mapped with the help of the Radial Basis Function Neural Network (RBFNN) architecture tuned by a novel meta-heuristic algorithm, namely, the Cooperative Search Optimisation Algorithm (CSOA). The architecture presented is able to automatically learn the kinematic properties of the manipulator. Learning is accomplished iteratively based only on the observation of the input–output relationship. Related simulations are carried out on a 3-Degrees of Freedom (DOF) manipulator on the Robot Operating System (ROS). The dataset created from the simulation is divided 65–35 for training–testing of the proposed model. The metrics used for model validation include spread value, cost and runtime for the training dataset, and Mean Relative Error, Normal Mean Square Error, and Mean Absolute Error for the testing dataset. A comparative analysis of the CSOA-RBFNN model is performed with an artificial neural network, support vector regression model, and with with other meta-heuristic RBFNN models, i.e., PSO-RBFNN and GWO-RBFNN, that show the effectiveness and superiority of the proposed technique.

show abstract

Kinematic Estimation with Neural Networks for Robotic Manipulators

Cited by 11 publications

References 10 publications

Experimental validation of a neuro-fuzzy controlled mobile manipulator for the inspection of automobile exhaust system

Experimental validation of a neuro-fuzzy controlled mobile manipulator for the inspection of automobile exhaust system

Transfer Learning-Based Artificial Neural Network for Forward Kinematic Estimation of 6-DOF Robot

Forward Kinematic Modelling with Radial Basis Function Neural Network Tuned with a Novel Meta-Heuristic Algorithm for Robotic Manipulators

Contact Info

Product

Resources

About