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

Traditional controllers have limitations as they rely on prior knowledgeabout the physics of the problem, require modeling of dynamics, and struggle toadapt to abnormal situations. Deep reinforcement learning has the potential toaddress these problems by learning optimal control policies through explorationin an environment. For safety-critical environments, it is impractical toexplore randomly, and replacing conventional controllers with black-box modelsis also undesirable. Also, it is expensive in continuous state and actionspaces, unless the search space is constrained. To address these challenges wepropose a specialized deep residual policy safe reinforcement learning with acycle of learning approach adapted for complex and continuous state-actionspaces. Residual policy learning allows learning a hybrid control architecturewhere the reinforcement learning agent acts in synchronous collaboration withthe conventional controller. The cycle of learning initiates the policy throughthe expert trajectory and guides the exploration around it. Further, thespecialization through the input-output hidden Markov model helps to optimizepolicy that lies within the region of interest (such as abnormality), where thereinforcement learning agent is required and is activated. The proposedsolution is validated on the Tennessee Eastman process control.

Specialized Deep Residual Policy Safe Reinforcement Learning-Based Controller for Complex and Continuous State-Action Spaces