抄録
on the operation for discharge-type micro-tunneling robots, heading control performance of the driving machine to track a planned trajectory largely depends upon the skill of operators and efforts toward realization of autonomous navigation system for the tunneling robots have been successively made. In this paper, the dynamic behavior of the driving machine of the robot for a head angle operation is analyzed and two types of heading control models, having motion constraint exerted from the soil surrounding the robot on the machine head tip and the rear end respectively, are introduced. Then Kalman filter is applied to estimate the posture of the driving machine and the model parameters matching the soil condition, for which direct observation is almost impossible. In the simulation using the driving data from an actual construction site, on-line estimation performances of the models are evaluated to show that the heading control model with the motion constraint on the rear end gives relatively good estimation performance and is expected to be useful to the design of heading controller for the tunneling robot.
