Estimation of Action/Reaction Forces for the Bilateral Control Using Kalman Filter

Abstract

As haptic technology has advanced in the past decade, many sensor devices have been built and commercialized for providing the sense of action/reaction forces or moment feedback to the human operator. In the ideal bilateral control of the haptic system, the force and position control should achieve with the same actions of force and motion movements. An external force should be sensed quickly after contact with an unknown environment. This paper presents formulation and application of a Kalman filtering technique for control of master-slave robots contacting environment in a bilateral control. The different types of motion sensors are mounted on each robot to provide the position and acceleration data in the horizontal dimension. A Kalman-filter-based state observer and a Kalman-filter-based disturbance observer have been designed to estimate action/reaction forces. This paper also deals with the construction of bilateral control with the use of data provided by the different types of motion sensors. Compared to the other conventional disturbance observer or state observer, the bilateral control system based on the proposed method offers the advantages of wider bandwidth, faster response, and free from sensor noise. Thus, it is possible to provide high transparency and good perception of the environmental stiffness. The experimental results are provided to illustrate the performance of the proposed algorithms.