Inverse kinematics learning for robotic arms with fewer degrees of freedom by modular neural network systems

Oyama, Eimei; Maeda, Takao; Gan, John Q.; Rosales, Eric M.; MacDorman, Karl F.; Tachi, Susumu; Agah, Arvin

doi:10.1109/iros.2005.1545084

Cited by 41 publications

(16 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As common in literature [4,10,7,17], the presented methods are evaluated on a simulated, planar robotic arm. The arm has three joints and the desired pose is the two dimensional position of the tip of the last link.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Rapid explorative direct inverse kinematics learning of relevant locations for active vision

Öfjäll

Felsberg

2013

2013 IEEE Workshop on Robot Vision (WORV)

View full text Add to dashboard Cite

An online method for rapidly learning the inverse kinematics of a redundant robotic arm is presented addressing the special requirements of active vision for visual inspection tasks. The system is initialized with a model covering a small area around the starting position, which is then incrementally extended by exploration. The number of motions during this process is minimized by only exploring configurations required for successful completion of the task at hand. The explored area is automatically extended online and on demand.To achieve this, state of the art methods for learning and numerical optimization are combined in a tight implementation where parts of the learned model, the Jacobians, are used during optimization, resulting in significant synergy effects. In a series of standard experiments, we show that the integrated method performs better than using both methods sequentially.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Different variations of neural networks have been popular approaches to this problem [11]. The results have been improved by using modular neural networks [9,10], where several different neural networks are trained and the output from the locally most suitable network is used.…”

Section: Previous Workmentioning

confidence: 99%

Rapid explorative direct inverse kinematics learning of relevant locations for active vision

Öfjäll

Felsberg

2013

2013 IEEE Workshop on Robot Vision (WORV)

View full text Add to dashboard Cite

show abstract

“…Since the desktop robotic arm uses a single‐chip microcomputer with limited performance as an arithmetic unit, which requires the computation of the inverse solution algorithm cannot be too complicated. The last three joints of the desktop robotic arm analyzed in this paper intersect at one point (the wrist point of the robotic arm), which satisfies the Pieper criterion , so the inverse kinematic solution of the robotic arm has closed‐form solutions. For a robotic arm with this structure, the first three joint angles determine the position of the wrist point, so they are solved by the geometric method, and then the last three joint angles are solved separately by the Z‐Y‐Z Euler angle inverse transform method.…”

Section: Algorithm—kinematicsmentioning

confidence: 99%

A teaching method for the theory and application of robot kinematics based on MATLAB and V‐REP

Zhou

Xie

Xuan

et al. 2019

Comp Applic In Engineering

View full text Add to dashboard Cite

With the continuous development of the world, robots are necessary for achieving high production rates and competitiveness. With an increasing robot demand, many universities offer robotics courses. These courses teach students the mathematical principles of robot‐related algorithms. However, it is difficult for students to apply pure mathematical algorithms in robotics to real robot development in a traditional classroom. This paper's motivation is to present a new robotics teaching method: Algorithm, Virtual experiment, Programming and Controller (AVPC), which shows students the steps of complete robot algorithm development, verification, and application process, taking the kinematics of an educational desktop 6‐DOF robotic arm as a typical teaching case. For the low‐cost educational desktop six‐axis robotic arm, an inverse kinematics algorithm is proposed. The inverse solution algorithm is realized in MATLAB, and then the dynamic model of the desktop robotic arm is established in Virtual Robot Experimentation Platform (V‐REP). The inverse kinematics algorithm in MATLAB is used to control the virtual robot arm movement in V‐REP to the specified position and orientation through the application programming interface (API) interface, thus verifying the correctness of the inverse kinematics algorithm. The inverse algorithm is implemented by using C language function in the microcontroller, and the running time of the algorithm is compared with that of the iterative solution method. In the present study, we investigated the students’ feelings and exam scores of using the new teaching method. The results show that the new teaching method has a significant impact on the improvement of students’ learning effect of robotics.

show abstract

“…Different variations of neural networks have been popular approaches to this problem [21]. The results have been improved by using modular neural networks [17,18], where several different neural networks are trained and the output from the locally most suitable network is used.…”

Section: Inverse Kinematicsmentioning

confidence: 99%

Online Learning of Vision-Based Robot Control during Autonomous Operation

Öfjäll

Felsberg

2015

Cognitive Systems Monographs

View full text Add to dashboard Cite

Online learning of vision-based robot control requires appropriate activation strategies during operation. In this chapter we present such a learning approach with applications to two areas of vision-based robot control. In the first setting, selfevaluation is possible for the learning system and the system autonomously switches to learning mode for producing the necessary training data by exploration. The other application is in a setting where external information is required for determining the correctness of an action. Therefore, an operator provides training data when required, leading to an automatic mode switch to online learning from demonstration. In experiments for the first setting, the system is able to autonomously learn the inverse kinematics of a robotic arm. We propose improvements producing more informative training data compared to random exploration. This reduces training time and limits learning to regions where the learnt mapping is used. The learnt region is extended autonomously on demand. In experiments for the second setting, we present an autonomous driving system learning a mapping from visual input to control signals, which is trained by manually steering the robot. After the initial training period, the system seamlessly continues autonomously. Manual control can be taken back at any time for providing additional training.

show abstract

Inverse kinematics learning for robotic arms with fewer degrees of freedom by modular neural network systems

Cited by 41 publications

References 17 publications

Rapid explorative direct inverse kinematics learning of relevant locations for active vision

Rapid explorative direct inverse kinematics learning of relevant locations for active vision

A teaching method for the theory and application of robot kinematics based on MATLAB and V‐REP

Online Learning of Vision-Based Robot Control during Autonomous Operation

Contact Info

Product

Resources

About