首页>
外国专利>
Method for computer-based learning of regulation and/or control of e.g. gas turbine, involves performing discretization by optimizing target, which is designed so that deviation of discrete hidden state from continuous hidden state is less
Method for computer-based learning of regulation and/or control of e.g. gas turbine, involves performing discretization by optimizing target, which is designed so that deviation of discrete hidden state from continuous hidden state is less
The method involves learning an action selection rule on an estimated hidden state with a learning and/or optimizing method e.g. prioritized sweeping method. An action to be performed at a technical system is delivered using the learnt action selection rule based on state vectors of the technical system. Discretization of the hidden state is performed by optimizing a discretization target (Tvq) in context of estimation of the hidden state. The discretization target is designed such that a deviation of a discrete hidden state (Sd) from a continuous hidden state is less. Independent claims are also included for the following: (1) a method for controlling and/or regulating a technical system (2) a computer program product comprising instructions for performing a method for computer-based learning of regulation and/or control of a technical system.
展开▼
