Objective Ring laser gyro is widely used in navigation, positioning, precision goniometer and other fields. Due to the backscattering of the reflectors and the non-uniformity of the optical loop, ring laser gyro has the lock-in phenomenon. In order to reduce the influence of the locking zone, the laser gyro must be biased. Mechanical dither bias is the method with the highest accuracy and is most widely used. However, mechanical dither bias has the defect that it needs to pass through the locking region twice in one cycle, and certain rotation signal loss will be generated each time it passes through the locking region. Jitter noise injection can randomize the rotation signal loss during the locking process, but it cannot eliminate this error, which will generate random walk error in the output of the gyroscope. In order to eliminate the lock-in error, the lock compensation is carried out.
Methods The lock compensation can obtain the error through the locking area and further compensate the error. In this paper, the lock compensation of laser gyro is realized for the first time through reasonable engineering design. The two instantaneous beat signals of the photocell are obtained by a high-speed ADC. After filtering the two signals, it can be judged whether the gyroscope has passed the locking area. If so, we can process it through the process in Fig.1. By orthogonal demodulation of read signal of laser gyro, the output pulse number of gyro can be obtained by reversible counter. Through the compensation expression, the compensation expression of the locking area is obtained.
Results and Discussions The gyro output signal without locking region compensation is shown (Fig.2), and the gyro output signal with locking region compensation according to the formula is shown (Fig.4). The comparison between the two figures shows that the output fluctuation of laser gyro after compensation is much smaller than that before compensation. The data in Fig.2 and Fig.4 were analyzed respectively by Allan variance, and the results were shown (Fig.5). It can be seen that the Allan variance of the data after compensation moved down much more than that before compensation. According to the data fitting of the two curves, the random walk before the lock compensation can be calculated as 1.53 \mathrme\text-3\left(^\circ\right)\sqrt \rmh , and the random walk after the lock compensation is 3.14\rme\text- 4 \left(^\circ\right)/ \sqrt \rmh , which is only 1/5 of the one before the compensation. It is confirmed that the lock compensation can reduce the random walk of the gyro indeed.
Conclusions The random error of each lock-in crossing in the ring laser gyro can generate the angle random walk (ARW) error in the output. In the frequency domain, the ARW error can be extended to the frequency band of useful signals, which is difficult to be filtered out by filtering method. Therefore, the random walk determines the ultimate accuracy of the navigation system. By recording the lock-in error of every lock crossing, the ARW of laser gyro is reduced. The Allan variance method is used to analyze the effect of lock compensation. The experimental results of a gyroscope show that the ARW after lock compensation is reduced to 1/5 of the original value. This is the first report of lock-in error compensation in engineering.
DownLoad: