CAJ | 학술논문

研制了一款基于磁阻传感器HMC1052/1051Z和MEMS加速度计ADXL203的磁罗盘，介绍了三轴磁阻罗盘的测量原理并分析了磁罗盘工作过程中可能存在的误差及其来源。针对样机的水平状态，利用前馈神经网络算法，建立了以测量航向角为输入、补偿后的航向角为输出的三层BP神经网络模型，并用样机的航向角采样数据进行仿真验证。实验数据表明，采用该BP神经网络补偿算法，可将航向角的精度从±35.73°提高至±0.5°以内，研制的样机可应用于普通导航领域。
연제료일관기우자조전감기HMC1052/1051Z화MEMS가속도계ADXL203적자라반，개소료삼축자조라반적측량원리병분석료자라반공작과정중가능존재적오차급기래원。침대양궤적수평상태，이용전궤신경망락산법，건립료이측량항향각위수입、보상후적항향각위수출적삼층BP신경망락모형，병용양궤적항향각채양수거진행방진험증。실험수거표명，채용해BP신경망락보상산법，가장항향각적정도종±35.73°제고지±0.5°이내，연제적양궤가응용우보통도항영역。
A magnetic compass based on magneto resistive sensor HMC1052/1051Z and MEMS accelerometer ADXL203 is designed, whose measuring principle is introduced and possible error and its source are analyzed. For the horizontal situation, according to the feedforward neural network, a 3 layers BP network is established, which input is measured course angle and output is compensated course angle, and the course angle sampling data of the compass system are used to simulate. The experiment results show that using the BP neural network compensate algorithm, the course angle precision can be improved from±35.73to±0.5or better, and the compass system can be used in common navigation field.