基于比例加速退化建模的单台机载设备剩余寿命自适应预测

针对现有机载设备剩余寿命(remaining useful lifetime, RUL)预测方法在新研单一样本条件下, 无法应用于加速退化试验场景的问题, 本文基于比例关系模型提出了一种加速退化场景下适用于单个试验样本的自适应RUL预测方法。首先, 依据加速退化环境下Wiener过程存在的漂移/扩散系数比例关系, 构建考虑设备个体差异与测量误差的非线性随机退化模型; 其次, 针对加速退化试验存在单一受试样本的情况, 提出了基于期望最大和卡尔曼滤波联合算法的参数自适应估计方法; 然后, 基于卡尔曼滤波原理在线更新目标设备的退化状态, 并推导出设备剩余寿命的概率密度函数; 最后, 通过对单台行波管加速退化实测数据进行分析, 验证了方法的正确性和优势。

Adaptive prediction of remaining useful lifetime for the single airborne equipment based on the proportional accelerated degradation modeling

Aiming at the problem that the existing prediction methods of remaining useful lifetime (RUL) for the airborne equipment cannot be applied to the accelerated degradation test scenarios under the condition of a single sample of the new research, this paper proposes an adaptive prediction method of RUL for the single equipment under the accelerated degradation scenario based on the proportional relationship model. Firstly, according to the ratio relationship of drift and diffusion coefficient in the Wiener process under accelerated degradation environment, a nonlinear stochastic degradation model considering individual differences and measurement errors is established. Secondly, for the situation that the accelerated degradation test has only a single sample, a parameter adaptive estimation method based on the joint algorithm of expectation maximum and Kalman filter is proposed. Then, the Kalman filter principle is used to online update the degradation state, and the probability density function of RUL is derived. Finally, through the analysis of the measured data for a single traveling wave tube under accelerated degradation test, the validity and advantages of this method are verified.