Multistage Bayesian fusion evaluation technique endorsing immersive virtual maintenance

Abstract

Immersive virtual maintenance is a standard practice for the evaluation of material sustainability. New techniques of service evaluation rely largely on multiple samples collected by realistic on-site tests, and it is complex for them to incorporate valuable knowledge from immersive virtual maintenance, leading to a high expense of maintenance evaluation. This paper therefore proposes a weighted data fusion approach using the t-test and F-test for virtual simulation data and real-life test data, based on a product's multi-stage immersive virtual maintenance simulation studies. The virtual and real fusion samples is then periodically fused by the multistage Bayesian iteration fusion method. Then to iteratively fuse the simulated and actual fusion samples, the multistage Bayesian iteration fusion method is used. Finally, on the basis of the maximum posterior statistical analysis of the final functional on-site samples, the maintenance index values are found. The results of the simulation show that the proposed approach can effectively combine virtual and real multi-stage maintenance details, reduce the total number of on-site maintenance test samples, and obtain more detailed results for maintenance evaluation.