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

Multi-agent reinforcement learning (MARL) has been applied and shown greatpotential in multi-intersections traffic signal control, where multiple agents,one for each intersection, must cooperate together to optimize traffic flow. Toencourage global cooperation, previous work partitions the traffic network intoseveral regions and learns policies for agents in a feudal structure. However,static network partition fails to adapt to dynamic traffic flow, which willchanges frequently over time. To address this, we propose a novel feudal MARLapproach with adaptive network partition. Specifically, we first partition thenetwork into several regions according to the traffic flow. To do this, wepropose two approaches: one is directly to use graph neural network (GNN) togenerate the network partition, and the other is to use Monte-Carlo tree search(MCTS) to find the best partition with criteria computed by GNN. Then, wedesign a variant of Qmix using GNN to handle various dimensions of input, givenby the dynamic network partition. Finally, we use a feudal hierarchy to manageagents in each partition and promote global cooperation. By doing so, agentsare able to adapt to the traffic flow as required in practice. We empiricallyevaluate our method both in a synthetic traffic grid and real-world trafficnetworks of three cities, widely used in the literature. Our experimentalresults confirm that our method can achieve better performance, in terms ofaverage travel time and queue length, than several leading methods for trafficsignal control.

Feudal Multi-Agent Reinforcement Learning with Adaptive Network Partition for Traffic Signal Control