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

Controller synthesis is in essence a case of model-based planning fornon-deterministic environments in which plans (actually ''strategies'') aremeant to preserve system goals indefinitely. In the case of supervisory controlenvironments are specified as the parallel composition of state machines andvalid strategies are required to be ''non-blocking'' (i.e., always enabling theenvironment to reach certain marked states) in addition to safe (i.e., keep thesystem within a safe zone). Recently, On-the-fly Directed Controller Synthesistechniques were proposed to avoid the exploration of the entire -andexponentially large-environment space, at the cost of non-maximalpermissiveness, to either find a strategy or conclude that there is none. Theincremental exploration of the plant is currently guided by adomain-independent human-designed heuristic. In this work, we propose a newmethod for obtaining heuristics based on Reinforcement Learning (RL). Thesynthesis algorithm is thus framed as an RL task with an unbounded action spaceand a modified version of DQN is used. With a simple and general set offeatures that abstracts both states and actions, we show that it is possible tolearn heuristics on small versions of a problem that generalize to the largerinstances, effectively doing zero-shot policy transfer. Our agents learn fromscratch in a highly partially observable RL task and outperform the existingheuristic overall, in instances unseen during training.

Exploration Policies for On-the-Fly Controller Synthesis: A Reinforcement Learning Approach