This thesis deals with causality of bilateral teleoperation systems. By performing the causality analysis, physical behaviors of teleoperation systems are well explained, and the analysis enables improvement in stability and transparency of bilateral teleoperation. Furthermore, control strategies for real teleoperation tasks are proposed based on the causality.
By causality, we mean dependency relationship; which is independent (cause) and which is dependent (effect). Causality is one of the physical characteristics of physical systems which involve energy interaction. Causality has been considered important for correct modeling, identification of system failure, and proper controller design. However, unfortunately, there are few works which consider causality of teleoperation systems where stability and transparency have been major issues.
A teleoperation system consists of a human operator, master arm, slave arm, and environment. These subsystems mutually have dependency relationship, i.e., causality, therefore causality analysis helps to understand physical behavior of teleoperation system and, furthermore, it can bring benefits for improving stability and transparency of teleoperation systems. This thesis is motivated in this context.
By using bond-graph modeling, the causalities in teleoperation systems are identified. Four models are proposed depending on the causalities of operator-master part and slave-environment part. After identifying four models, the benefits of causality-based approach are determined in terms of stability and transparency.
First, for stability, causality-based stability provides broader and easier design of stable control architectures compared to previous stability criterion. In previous works, absolute stability has been adopted for stable teleoperation system design. However, absolute stability has difficulty even for small time delay between master and slave, and it has been known that absolute stability compromises transparenc...