On application of cumulative sums algorithm to recurrent estimation of parameters of non-stationary control objects

Recurrent least squares method is supplemented with cumulative sums to detect abrupt changes in the parameters of the non-stationary object. The proposed approach to the estimation of parameters of the object is tested on a model example, and the experimentally obtained dependence of the liquid level in the reservoir, which has a piecewise behavior. The proposed approach improves the speed of adaptation of estimates to abrupt changes in the parameters of the object.