This study proposes three efficient algorithms for online computing of fingertip force modeling and noise involved in 1-D, 2-D and 3-D models. Fast and effective real-time detection for fingertip force modeling is extremely important for successful manipulation on an object by a robot hand in a real situation. Meanwhile, signal processing for a desired force signal is necessary if unwanted signals and noise cannot be ignored in practical applications. Butterworth low-pass filters designed in this study can be effectively used for filtering noise and obtaining flatter and smoother force modeling detection signals. An EtherCAT slave circuit is designed and force detection models using Beckhoff TwinCAT 3.1 provide feedback signals. The calculation is made during several real-time experiments.

In this study a novel Microcontroller unit (MCU) circuit based on FPGA for EtherCAT system is presented. The resource utilization statistics of the MCU circuit are provided and the performance of the MCU circuit is analyzed. The first objective is to understand the feasibility of the approach, i.e., whether it is possible for the MCU to drive the EtherCAT slave interface, be reasonable for real-time performance and be stable at different communication SPI frequencies. The second objective is to give the MCU circuit developers some valuable guidelines. Furthermore, it is verified that the proposed MCU circuit based on FPGA is stable, reliable and real-time tested using the online debugging tool SignalTap II.

A novel strategy of online force distribution for multi-finger robot hand is developed based on impedance control. The real-time controller system including the finger dynamics and multi-fingers constraints is considered as a nonlinear system. Because when the robot hand grasps the object, the fingertip force is usually changed which could cause the grasp unstable. Our method gains the ability to deal with online grasping force distribution and can satisfy the real-time requirements at the same time. Experiment shows that our method provides the most feasible grasp configurations in real-time performance.

A multi-finger dynamics model has been presented in this study, which contains a single finger dynamics model equation and a restraint equation between fingers based on Lagrangian multiplier controller. To validate the model, an EtherCAT master and slave platform has been developed based on FPGA. Meanwhile, the multi-finger dynamics algorithm has been designed in the TwinCAT. Finally, the experiments demonstrate this strategy can be implemented and operated by online grasping object.

According to the EtherCAT process data interface networks system designs, we determined that the Synchronous and Asynchronous μController scheme evaluations based on the FPGA state machine algorithm. The present study focuses on the Schematic of μController interconnection between FPGA and ESC based on the VHDL state machine algorithm, which includes state machine programs. In addition, the performance of the Synchronous and Asynchronous μController was verified through numerical simulations. Furthermore, we found the simulations gave further evidence that the 16 bits synchronous and Asynchronous μController signals can satisfy the two write access and a read access requirements for the EtherCAT networks communication. Finally, the results include the positive and negative aspects of the Synchronous and Asynchronous μController networks communication, and the comparison the results of the Synchronous, Asynchronous and SPI communication evaluations done by the researchers. © 2013. WSEAS Transactions on Communications.

The objective of this study is to show the applications of the simulation implementation of QNX identified model combining with Matlab/Simulink platform, QNX C/C++ compiler and QNX real-time operating system. The HIT/DLR Hand II dexterous hand QNX real-time simulation model is generated by the Simulink platform and QNX C/C++ compiler link files. Moreover, this model can be implemented in the QNX real-time system. Simulations in the QNX real-time system show that the HIT/DLR Hand II dexterous hand QNX model which based on the methods in this research, was able to offer an attractive performance during both real-time simulation and feedback operations.