Path generation and optimization for DBB measurement with continuous data capture

Abstract

In this paper we present a systematic approach to generating a path for continuous Double Ball Bar(DBB) measurement. The generated path is feasible, free of collision, jerk-optimized and sensitive to the perturbation in kinematic parameters. The approach could be automated and be applied not only to parallel kinematic machines(PKM) but also to serial mechanism. We will discuss about the properties of continuous DBB measurement and conclude that the condition number of identification Jacobian is a proper index to describe the sensitivity of a continuous path to the perturbation in kinematic parameters. The sensitivity of the generated path will be experimentally validated on a Stewart-Gough platform with comparison to two otherwise designed paths. Measurement shows that the generated path is also suitable for time-demanding tasks.