Intelligent Controller for Robotic Motion Control

Abstract

A self-organizing fuzzy controller (SOFC) has been developed to control complicated and nonlinear systems. However, it is arduous to choose an appropriate learning rate and a suitable weighting distribution of the SOFC to achieve satisfactory performance for system control. Furthermore, the SOFC is mainly used to control single-input single-output systems. When the SOFC is applied to manipulating a robotic system, which is an example of multiple-input multiple-output systems, it is difficult to eliminate the dynamic coupling effects between the degrees of freedom (DOFs) of the robotic system. To address the problems, this study developed a self-organizing fuzzy radial basis-function neural-network (RBFN) controller (SFRBNC) for robotic systems. The SFRBNC uses an RBFN to regulate in real time these parameters of the SOFC to optimal values, thereby solving the problem faced when the SOFC is applied. The RBFN has coupling weighting regulation ability, so it can eliminate the dynamic coupling effects between the DOFs for robotic system control. From the experimental results of the 6-DOF robot tests, the SFRBNC demonstrated better control performance than the SOFC.Index Terms-Radial basis-function neural-network (RBFN), robotic systems, self-organizing fuzzy controller (SOFC).

Development of a Virtual Platform for Telepresence Control of an Underwater Manipulator Mounted on a Submersible Vehicle

“…Therefore, the virtual platform provides a very good training environment for the operators. In addition, the virtual platform can be used to test and validate motion control methods for operating the submersible vehicle and manipulator [26,27]. It can also be used to supervise a real underwater task where the developers do not have a direct view of the system.…”

Research on Control Method Based on Real-Time Operational Reliability Evaluation for Space Manipulator

“…Zhang et al [6] aimed at trajectory tracking of robot manipulators without velocity information; the study proposed an output feedback PD control without measuring joint velocities dispense with model. Lian [15] developed a self-organizing fuzzy radial basis-function neural network (RBFN) controller (SFRBNC) for robotic systems.…”

Fuzzy-neural-network control for robot manipulator via sliding-mode design

“…In the literature of robotic control, most of the intelligent schemes combine adaptive control strategy [7]- [11]. Neural network (NN) has an inherent ability to learn and approximate a nonlinear function, which is utilized in the robotic controllers [7]- [9] to model complex processes, to compensate for unstructured uncertainties and even to prevent actuator saturation [7] without a prior knowledge of system parameters.…”

“…Alternatively, fuzzy control comes from human expert experience rather than from mathematical models. Generally, fuzzy logic is employed in robot manipulator controllers to provide human logical thinking capabilities in order to obtain good control performance over uncertainties [10], [11]. How to build suitable fuzzy rules and guarantee the system stability is a challenge problem to be solved.…”