A frequency-domain stochastic stability criterion for human-controlled nonlinear systems [nuclear power plant control]

A class of human controlled systems is modelled as randomly-sampled nonlinear control systems. The transfer function of human controlled systems is defined and used to derive a criterion for the stability with probability 1' which can be considered as a stochastic analogue of the well-known circle criterion available for related classical deterministic feedback systems. In order to demonstrate the merits of the proposed criterion, a completely worked out example is given of the control of a nuclear power plant.