A systematic solution to precision control of modular robot manipulators without using joint torque sensing is presented in this paper for the first time. Using the virtual decomposition control (VDC) approach with embedded field programmable gate array (FPGA) logic devices, the proposed solution solves a long-standing problem of lacking control precision fundamentally associated with the modular robot manipulators. As a result, this solution allows modular robot manipulators to possess not only their traditional advantages (such as reconfigurability, flexibility, versatility, and ease of use) but precision control capability as well. A hierarchical master-slave control structure is used, which is supported by a high-speed communication system modified from SpaceWire (IEEE 1355), transferring a limited amount of data between the master and slave nodes at a rate of 1000 Hz. In each module, the FPGA logic implementation uses multiple sampling periods of 163.8 μs, 1.28 μs, and 20 ns. A gravity counterbalance spring provides a design option for the purpose of energy saving. Experimental results demonstrate unprecedented control precision, which is attributed to the use of both the VDC approach and embedded FPGA implementation. The ratio of the maximum position tracking error to the maximum velocity reaches 0.00012 s-more than an order of magnitude better than available technologies in control of robots with harmonic drives. The solution presented in this paper is also applicable to integrated robot manipulators using embedded FPGA controllers.

Modular (re-configurable) robots have been studied and developed over two decades. Compared to common industrial robot manipulators, modular robots usually end up showing relatively poorer control performance. This is mainly because most of modular robots' designs typically concentrate on mechatronic interfaces, packaging and re-configurable features, leaving little space for dynamics and control considerations. In our current research frame, we are looking at "virtual decomposition control" (VDC) to significantly improve modular robots dynamics and control performance. A key element to successful implementation of the related algorithms has to do with the inter-modules communication link capabilities. Presented here is the first 2 DOF prototype built to demonstrate the feasibility of implementing VDC on a real system, including the Spacewire Bus based communication protocol used.

Modular or re-configurable robots have been studied and developed over two decades. Most researches focus on mechatronic interfaces and re-configurable capabilities. However, less attention has been paid to dynamics and control. Consequently, the control performance of a modular robot has never been comparable with an integrated robot, due to the lack of proper handling of the dynamic interactions among the modules. In this paper, the application of the virtual decomposition control to modular robot manipulators is discussed. A high-speed databus with a data rate of 100 Mbps is used for necessary information exchange among the modules. The dynamics based control is fully handled by the local embedded controllers, whereas the host computer handles the kinematics related computation. The stability of the entire robot is rigorously guaranteed. This research aims at giving the modular robots the comparable control performance as the integrated robots, while keeping the fundamental feasibilities such as low cost for mass production, high flexibility, and easy use and expansion.

Velocity estimation is crucial to certain robotic applications involving high bandwidth modeling and control. In the conventional approaches, the velocities generated from encoders or tachometers are quite noisy and low-pass filters are usually engaged to generate usable velocity signals. The low-pass filter, however, cause non negligible phase lag that may severely affect both modeling and control accuracy in the middle and high frequency range. In this paper, two approaches of using a fusion of an encoder and an imperfect accelerometer are proposed to estimate accurate velocities. The two approaches, namely the two-channel approach and the observer-based approach, estimate velocities by using proper frequency weightings over the encoder and accelerometer signals. The encoder mainly contributes to the low-frequency part of the velocity estimation and the accelerometer mainly contribute to the high-frequency part of the velocity estimation. An adaptive mechanism for estimating the accelerometer gain is also presented. The effectiveness of the two proposed velocity estimation approaches is verified experimentally with respect to a one degree-of-freedom robot in terms of both rigid contact modeling and control