Computational approach for deriving optimal design laws from the data. (IMAGE)
Caption
The central is to leverage the sampled data to extract laws from data. To this end, pre-process the actions and outputs of the dynamical system and construct the feedforward signals that will be used for the feedback gain learning and the design of an online real-time control loop. Then, measure the input-output data, as well as the feedforward signals, over discrete-time series, based on which the discrete-time data samples are assembled using the tensor product. Calculate the Bellman equation for optimality via policy iterations. Through policy evaluation and improvement, the optimal feedback gain is obtained from the discrete-time data samples with rigorous mathematical operations and convergence deduction. Finally, both the feedforward signal and the feedback gain contribute to the optimal decision law.
Credit
©Science China Press
Usage Restrictions
Use with credit.
License
Original content