论文部分内容阅读
光纤陀螺具有较高实际应用价值,其性能与温度变化密切相关,导致测量系统的测量精度不稳定,针对光纤陀螺的非线性变化特性,提出一种小波神经网络的光纤陀螺温度漂移误差补偿模型。首先收集光纤陀螺温度漂移误差与光纤陀螺输出之间历史样本,然后采用神经网络对它们的关系进行非线性逼近,并引入小波基函数克服传统神经网络泛化能力差的缺陷,建立光纤陀螺温度漂移误差补偿模型。实验结果表明,本文光纤陀螺温度漂移误差补偿模型具有简单、易实现等优点,而且性能要其于其它类型的神经网络,补偿精度得到不同程度的提高。
FOG has high practical value, and its performance is closely related to temperature changes, which leads to the instability of the measurement accuracy of the measurement system. Aiming at the nonlinear variation characteristics of FOG, a novel WFN temperature drift error compensation model is proposed. Firstly, we collected the historical samples between the temperature drift error of the FOG and the output of the FOG, then used the neural network to approximate the relationship between them and introduced the wavelet basis function to overcome the defect of the poor generalization ability of the traditional neural network, established the FOG temperature drift Error compensation model. The experimental results show that the temperature drift error compensation model of the fiber optic gyroscope is simple and easy to implement, and the performance is better than other types of neural networks, the compensation accuracy is improved to some extent.