Fitness Functions for Detection, Selection an Adaptation in Stochastic Environments

I. Semoushin, M. Federova, and M. Sunoplya (Russia)


Fault detection, adaptive and intelligent control, stochasticsystems.


In a stochastic environment, it should be possible to handle stochastic fitness functions that cannot be strictly evaluated. This paper proposes and experimentally examines two dif ferent kinds of fitness functions used to detect change point and select the best performer from a filter population with the aim to identify the steady-state Kalman filter both in open- and closed-loop stochastic control environments.

Important Links:

Go Back