Analysis of Multi-sensor Attitude Measurement System on TBM

It is well known that the real-time high accuracy attitude measurement is required during the tunnel construction. The vibration is caused by hard rock tunneling, which makes the inclinometer fail to work, and then the current system can not meet the demand of auto-guide control. In this paper, we describe the fusion and test of an optic fiber gyroscope and inclinometer based on multi-sensor fusion algorithm. The proposed multi-sensor is used for measuring the attitude of TBM(Tunnel Boring Machine),which can solve those problems. Inclinometer has the advantage of high accuracy output and gyro is not sensitive to vibrations. The system discusses fuse information from two sensors and develops the fusion algorithms. Firstly, stochastic error model of gyro is built on the basis of ARMA(auto regressive moving average) modeling. Secondly, the detail modeling of fuse information for multi-sensor is made by using Kalman filter, which estimates and compensates zero drift of gyro with the high accurate output of inclinometer. The attitude under the strong vibration is calculated by compensated gyro. Finally, using test results, it is shown that the multi-sensor information fusion algorithm is effective and improves the accuracy of attitude measurement. It ensures that the error of multi-sensor attitude measurement system is less than 2 milliradian under the static and dynamic state.