PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 8, 2026Applied Sciences1 citationsOpen Access

Intelligent Scheduling Method for Bus Passenger Flow Prediction Using GCN-AM-LSTM

Research on an Intelligent Scheduling Method Based on GCN-AM-LSTM for Bus Passenger Flow Prediction

View Full Paper
Ask AI
Bookmark
Share

Authors

XJXiaolei JiZLZhe LiZGZhiwei Guo

Discussion

Loading...

Member takes

Overview

Prediction model reduces Mean Absolute Error by 14% in bus passenger flow, suggesting more efficient scheduling solutions.

Key Points

  • This research aims to enhance passenger flow prediction accuracy and scheduling efficiency in public transit systems.
  • Analyzed passenger flow variation patterns spatially and temporally.
  • Constructed spatiotemporal matrices and applied dimensionality reduction methods.
  • Developed a prediction model using GCN-AM-LSTM and a dynamic scheduling strategy.
  • Achieved a 14% reduction in Mean Absolute Error compared to CNN and LSTM models for predictions.
  • Decreased the number of departures by 15.24% and reduced passenger waiting time costs by 3.7%.
  • Transit operating costs lowered by 3.19%, balancing service quality and operational efficiency.

Cite This Study

Ji et al. (2026) studied this question.

synapsesocial.com/papers/69acc58f32b0ef16a404fd7ehttps://doi.org/10.3390/app16052525
View Full Paper
Ask AI
Bookmark
Share