Tatsuya Doi, Hironao Yamada, Toshihisa Ikemoto, and Hiroshi Naratani

Abstract:
Simulation of a pneumatic hand crane for work support was developed for system design. The crane's mathematical model consists of a pneumatic cylinder model, a crane link model, and a controller that includes a human model to passively operate the hand crane by operating force. The human model's inverse system learns using a neural network. The feasibility of the simulation was demonstrated in comparison with measurement results from operating force and acceleration.

Keywords: power-assist, simulation, pneumatic, human model, neural network