Search

Identifying the actual time of a change brings a decrease in time range of searching for assignable causes leading to less cost. In this paper, the maximum likelihood approach is developed to estimate sporadic change point for the mean of a polynomial profile in Phase II which has not been performed yet in the literature. Estimation of the process parameters for the samples after the change point is carried out using filtering and smoothing estimation methods of dynamic linear models. The proposed procedures are applied after receiving an out-of-control signal from T2 control chart. The performance of the proposed change point estimators is also compared to the step and drift estimators' performance under sporadic change in the process mean. Simulation results confirm the effectiveness of the proposed methods in estimating sporadic change point.