Suboptimal robot motion for inexperienced situations through intelligent composite motion control: Method of knowledge array

Abstract

The intelligent composite motion control (ICMC) is a methodology to build up robot systems in which robots can realize complex and dexterous behaviours autonomously and adaptively by parameter optimization and utilization of experiential knowledge only if the motion control for fundamental element motions is given. In this paper, after reviewing the ICMC method of a knowledge array is presented, this is a method for realizing suboptimal motions for inexperienced situations by the use of experiential knowledge, in which the experiential knowledge is stored in the form of a knowledge array and is utilized to estimate suboptimal motions for inexperienced situations. The array is the generalization of the vector-matrix and it can easily and concisely express and perform such operations that are very difficult or impossible to be expressed in the conventional vector-matrix formulation. Therefore it is expected that the array will realize highly intelligent knowledge operations such as abstraction and broad interpretation. In the ICMC, all the robot behaviours from simple point to point (PTP) motions to complicated tasks are treated in a unified manner; thus the ICMC can be applied to current industrial robots as well as to future, much more complicated robot systems. It is shown that the method is effectively applied to the dextrous handling problem of flexible plates.