Control of the probability distribution of a system state based on its structure indicator

Аннотация

A nonlinear dynamic system with a random jump structure is considered. When the system’s structure changes, the plant’s nonlinear characteristic changes in a jump and the input random binary signal also changes. The structure change is registered by the structure indicator; however, its indications are error prone. A Bayesian recurrent algorithm for controlling the plant structure and probability distribution of the plant state is developed.