Random coefficient regression model residual life prediction method considering dependency time error

The invention discloses a random coefficient regression model residual life prediction method considering a dependency time error. The method comprises the following steps: A, establishing a high-reliability equipment performance degradation model under the condition that an error variance is reduced along with time; b, performing off-line unbiased estimation on prior parameters of the model; c, on-line updating the random coefficient based on the Bayesian theory; and D, predicting the residual life of the equipment. The invention provides the residual life prediction method of the random coefficient regression model considering the dependency time error, which not only can carry out fitting analysis on the performance degradation characteristic of the measurement uncertainty of the high-reliability equipment along with the reduction of time, but also can effectively predict the residual life of the equipment. Theoretical basis and technical support are provided for condition-based maintenance and decision ma