English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Bus timetable optimization is a key issue to reduce operational cost of buscompanies and improve the service quality. Existing methods use exact orheuristic algorithms to optimize the timetable in an offline manner. Inpractice, the passenger flow may change significantly over time. Timetablesdetermined in offline cannot adjust the departure interval to satisfy thechanged passenger flow. Aiming at improving the online performance of bustimetable, we propose a Deep Reinforcement Learning based bus Timetable dynamicOptimization method (DRL-TO). In this method, the timetable optimization isconsidered as a sequential decision problem. A Deep Q-Network (DQN) is employedas the decision model to determine whether to dispatch a bus service duringeach minute of the service period. Therefore, the departure intervals of busservices are determined in real time in accordance with passenger demand. Weidentify several new and useful state features for the DQN, including the loadfactor, carrying capacity utilization rate, and the number of strandingpassengers. Taking into account both the interests of the bus company andpassengers, a reward function is designed, which includes the indicators offull load rate, empty load rate, passengers' waiting time, and the number ofstranding passengers. Building on an existing method for calculating thecarrying capacity, we develop a new technique to enhance the matching degree ateach bus station. Experiments demonstrate that compared with the timetablegenerated by the state-of-the-art bus timetable optimization approach based ona memetic algorithm (BTOA-MA), Genetic Algorithm (GA) and the manual method,DRL-TO can dynamically determine the departure intervals based on the real-timepassenger flow, saving 8$\%$ of vehicles and reducing 17$\%$ of passengers'waiting time on average.

Deep Reinforcement Learning based Dynamic Optimization of Bus Timetable