Track inspection is dedicated to detecting track geometry-related potential defects which require several visits to accomplish the ob- jective due to uncertainties in a track deterioration model. Through updating belief about prior specifications of the model parameters using freshly available track inspection data, the current inspec- tion design would gain an improvement in terms of the value of investment (not necessary concerning cost reduction) for next main- tenance cycles. In the sense of maximising an opportunity of having uncertainty reduction, this study proposes a statistical-based method blending with information theory to assign a priority index to a discrete-time point (called a candidate) in a given time interval. High priority is, in general, dedicated for a candidate that expect- edly offer large cut in uncertainties i.e. high informativeness time for inspection. To lowering a false positive rate in the index assign- ment bearings and size of the ratio of adjustment are sequentially analysed before the final ranking gets to publish. Analysis of the simulation results corresponds to various settings of linear geome- try deterioration model establishes correlations between covariance and priority index convincingly. Also, the relationship between the time gap between consecutive inspections and a prior deterioration rate is found significant in this context. A detailed description of the proposed model development is presented in this paper.
|Title of host publication||The 9th International Conference on Computer Engineering and Mathematical Sciences 2020 (ICCSCM 2020)|
|Publication status||Published - 9 Jul 2020|
Bibliographical noteThe 9th International Conference on Computer Engineering and Mathematical Sciences 2020 (ICCSCM 2020), July 9-10, Langkawi, Malaysia.
- Uncertainty propagation
- track quality
- rail track inspection
- risk analysis