Grouping of Bus Stops for Aggregation of Route-Level Passenger Origin–Destination Flow Matrices

Abstract

Route-level, bus passenger origin–destination (O-D) matrices summarize useful information on travel patterns for use in route planning, design, and operations. However, the size of stop-to-stop O-D matrices makes it difficult to synthesize important flow patterns and to estimate stop-to-stop O-D passenger flows accurately. To reduce the size of the O-D matrix for improved estimation, analysis, and communication, stops can be grouped so that passenger flows between far fewer stop group pairs can be estimated and represented. The problem of grouping of bus stops for aggregation of passenger O-D flows is presented with the explicit objective of reducing the passenger O-D flow matrix while capturing important O-D flow characteristics. Two computationally efficient heuristic algorithms for solution of this problem are proposed. A set of empirical studies—conducted with automatic passenger counter data collected on major bus routes operated by the Ohio State University's Campus Area Bus System, the Central Ohio Transit Authority, and the Los Angeles County, California, Metropolitan Transit Authority—shows that the stop group configurations identified with the proposed heuristic algorithms are close to optimal and capture pertinent flow patterns and dominant clusters of stops much better than stop group configurations produced from land use characteristics. The heuristic algorithms can identify solutions for all possible numbers of stops with a reasonable computational time, whereas optimal configurations can be found for only a few groups on long routes.

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 — 11 citations exist upstream.

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