论文部分内容阅读
总方差方法被引入到光纤陀螺随机误差特性分析当中,有效克服了传统Allan方差法在长相关时间上计算易出现“崩溃”的问题。但用它直接分析光纤陀螺随机噪声存在算法偏差,不能真实反映角度随机游走噪声、量化噪声和指数相关噪声方差值,因此在算法上对总方差提出改进,使其适用于光纤陀螺噪声分析。对模拟的各项随机噪声进行仿真计算,验证改进后总方差方法可有效辨识噪声类型和水平。对光纤陀螺实测数据方差分析表明改进的总方差方法在平均因子较大的情况下可提高估计置信度,方差值稳定性好,比Allan方差法能更精确地分析出噪声项系数。
The total variance method is introduced into the analysis of random error characteristics of fiber optic gyroscope (FOG), which effectively overcomes the problem of “collapse” easily calculated by the traditional Allan variance method in long correlation time. However, it directly analyzes the deviation of the stochastic noise of the fiber optic gyroscope, and can not truly reflect the variance of random walk noise, quantization noise and exponential correlation noise. Therefore, the total variance of the algorithm is improved so that it can be applied to the fiber optic gyroscope noise analysis . Simulation of the random noise of each simulation to verify the improved total variance method can effectively identify the type and level of noise. The analysis of variance of measured data of FOG shows that the improved total variance method can improve the confidence of estimation under the condition of large average factor, and the stability of variance is good. The noise coefficient can be more accurately analyzed than Allan variance method.