2019 Volume 60 Issue 3 Pages 172-177
One of the technical problems encountered when dispatching trains using computation is how to reflect all the conditions or criteria that influence a decision that form part of the tacit knowledge that train dispatchers have acquired from experience in conventional rescheduling systems. A method has been developed from operation performance records, to generate decisions that would have been made by dispatchers with a high probability under specific conditions, namely operational dispatching rules (ODRs). A train rescheduling algorithm was created based on the mechanisms related to ODRs and its validity by comparing actual operation planning results with computed outputs.