Transport risk modeling for hazardous chemical transport Companies—A case study in China

Abstract

Traffic accidents of hazardous chemical transport vehicles strongly correlate with the operation and management level of road transport companies. An accurate risk assessment of these transport companies will play a critical role in improving their management and supervision and in turn the overall safety of roadways, property, and most importantly people's lives. Therefore, this study constructs a logistic regression scorecard model to evaluate transport risk of hazardous chemical transport companies and evaluates it using a case study in China. This study first selected 16 indicators from the dimensions of driver behaviors, driving performance, dangerous goods and company business operation to construct company user portraits. Next, a K-means++ algorithm was used to cluster the data samples of the companies on a monthly basis. On this basis, a scorecard was constructed based on a logistic regression scorecard model to realize and visualize the monthly risk portrait of companies. The constructed scorecard predicted transport risk of companies accurately. The results show that the more complex the types of dangerous goods transported, the higher the risk value of the company in the dangerous goods index. Moreover, for transport companies, the driver behaviors and driving performance on their trips are closely related to their business risk. In a practical application, the proposed scorecard can realize the dynamic risk monitoring of transport companies and enable managers and supervision departments to clarify where the risk is from. The transport companies can also carry out safety training and rectification for drivers and operations to reduce the occurrence of hazardous materials-related traffic accidents.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 14 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.