eduzhai > Applied Sciences > Engineering >

Off Policy Risk Sensitive Reinforcement Learning Based Optimal Tracking Control with Prescribe Performances

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: An off policy reinforcement learning based control strategy is developed forthe optimal tracking control problem to achieve the prescribed performance offull states during the learning process. The optimal tracking control problemis converted as an optimal regulation problem based on an auxiliary system. Therequirements of prescribed performances are transformed into constraintsatisfaction problems that are dealt with by risk sensitive state penalty termsunder an optimization framework. To get approximated solutions of the HamiltonJacobi Bellman equation, an off policy adaptive critic learning architecture isdeveloped by using current data and experience data together. By usingexperience data, the proposed weight estimation update law of the criticlearning agent guarantees weight convergence to the actual value. Thistechnique enjoys practicability comparing with common methods that need toincorporate external signals to satisfy the persistence of excitation conditionfor weight convergence. The proofs of stability and weight convergence of theclosed loop system are provided. Simulation results reveal the validity of theproposed off policy risk sensitive reinforcement learning based controlstrategy.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...