CAJ | 학술논문

　　根据快反镜中位移传感器的位置特点，利用ICA方法对传感器信号进行去噪。ICA算法的缺点是需要计算高阶统计量，计算量较大，会影响实时性能，而借助高性能计算硬件则可以弥补这方面的不足。仿真结果表明，经ICA去噪后，信噪比能提升15 dB左右，用该方法对采集的传感器信号去噪，信噪比增量在13~16 dB之间。可见，ICA方法能够有效地提升信号的信噪比，具有较好的实用性。
　　근거쾌반경중위이전감기적위치특점，이용ICA방법대전감기신호진행거조。ICA산법적결점시수요계산고계통계량，계산량교대，회영향실시성능，이차조고성능계산경건칙가이미보저방면적불족。방진결과표명，경ICA거조후，신조비능제승15 dB좌우，용해방법대채집적전감기신호거조，신조비증량재13~16 dB지간。가견，ICA방법능구유효지제승신호적신조비，구유교호적실용성。
The independent component analysis(ICD)algorithm is applied to denoising of the sensor signals according to the position feature of fast steering mirror sensors. The shortages of ICA algorithm are that it has to calculate the high?order sta?tistics and the mass computation affects the real?time preformance. These problems can be solved by high?capability computation hardware. The simulation results show that the signal?to?noise ratio can be increased by 15 dB by ICA denoising,and the incre?mental quantity is between 13 dB and 16 dB when this method is used to denoise eddy current sensor signals. Therefore,the ICA method can improve the signal?to?noise ratio effectively.