基于改进OPTICS算法的磁罗盘椭圆拟合误差补偿方法

In order to avoid the adverse effect of singular error points in the original data of magnetic compass on the accuracy of ellipse fitting, the ordering points to identify the clustering structure (OPTICS) algorithm based on density clustering is introduced into the conventional least squares ellipse fitting algorithm, and the OPTICS algorithm is improved to reduce the complexity and calculation time. Before ellipse fitting, the improved OPTICS algorithm is used to classify the original sensor data, identify and eliminate the singular error points in the data to prevent them from polluting the original data set and improve the fitting accuracy of ellipse fitting algorithm, so as to improve the calculation accuracy of heading angle. The experimental results show that compared with the conventional ellipse fitting algorithm, the root mean square of the total heading angle error in the four directions of east, south, west and north is reduced by 27.5%, and the calculation time of the improved OPTICS algorithm is 43.6% lower than that before the improvement.