Constant costate iterations for finite-horizon optimal control with nonlinear dynamics.
Lorenzo TarantinoMario SassanoSergio GaleaniAlessandro AstolfiPublished in: CDC (2022)
Keyphrases
- optimal control
- infinite horizon
- finite horizon
- nonlinear dynamics
- average cost
- dynamic programming
- feedback control
- neural network
- single item
- stochastic demand
- dynamical systems
- inventory models
- brownian motion
- inventory control
- control strategy
- production planning
- single product
- reinforcement learning
- optimal control problems
- markov decision process
- control law
- markov decision processes
- optimal policy
- optimal solution
- lost sales
- fixed cost
- multistage
- initial state
- lyapunov function