A novel tracking estimation algorithm for the detection beam's roll angle

Chunjun Chen,Xiaoyu Liu,Lu Yang,Ji Deng
DOI: https://doi.org/10.1007/s40430-024-05128-x
IF: 2.361
2024-08-31
Journal of the Brazilian Society of Mechanical Sciences and Engineering
Abstract:The detection beam is the carrier of the track irregularity detection system. The high-precision estimation of the detection beam's roll angle is the key technology to improve the detection accuracy of the cross-level irregularity and other items. Nevertheless, the traditional complementary filtering method has the drawbacks of low model approximation and an ambiguous cut-off frequency. To solve the above problems, a measuring system is designed on the inclinometer and the gyroscope to realize the high-precision tracking estimation of the detection beam's roll angle. The sensor error model and the angular motion model of the detection beam are combined to establish a large system model. Then a novel strong maneuvering nonlinear tracking algorithm (SMNT) for the detection beam's roll angle is proposed. A multi-body dynamic model is created in Simpack to replicate the angular motion of the detection beam. The simulation data are superimposed on the sensor noise to the SMNT algorithm. The simulation results show that the SMNT algorithm is the best among the four algorithms by considering convergence speed, estimation error, and robustness. The SMNT algorithm converges after 25 m, the maximum root mean square error is 0.0029° and the maximum cross-level irregularity error is 0.1028 mm. This SMNT algorithm can provide support for improving the detection accuracy of track irregularity parameters.
engineering, mechanical
What problem does this paper attempt to address?