Study on the compensation for a quartz accelerometer based on a wavelet neural network

In the position and azimuth determination system (PADS), the bias and scale factor of a quartz flexible accelerometer exhibit a non-ignorable drift after a quick turn-on and before the thermal balance is reached inside the system at different temperatures. To solve this problem, this paper presents a system compensation scheme. The multi-point temperature sensing experiment was repeated based on the realization of the hardware platform, and a wavelet neural network was used for the temperature compensation. Studies indicated that theselection of the temperature sensing points is one of the keys to achieve an effective compensation. Results show that the compensation scheme and algorithm increase the efficiency and reduce the preparation time; meanwhile the attitude precision can be improved to 8 arcsec, which meets the specification of the system.