Application of configurable models on Laguerre filters for finding parametric defects in continuous ACS

Bloshchinskiy Vladislav Dmitrievich, Shalobanov Sergey Victorovich, Shalobanov Sergey Sergeevich

The paper considers an algorithm using configurable diagnostic models based on Laguerre filters with only a recursive part for finding parametric defects in continuous automatic control systems. A comparison is made of two approaches to configuring models: combined and individual, in relation to the choice of model parameters for each control point in the diagnostic object. Numerical experiments were carried out to configure the diagnostic models, select sets of vectors of reference defect technical states of the diagnostic object, and diagnose introduced defects by calculating the distinguishability by diagnostic features. The accuracy of the algorithm as a whole and the effectiveness of configuring approaches for various diagnostic object are estimated. All numerical experiments and simulations were carried out in a software package developed in the Scilab/Xcos environment.

Keywords:parametric defect, automatic control system, diagnostic object, diagnostic feature, Laguerre filter, configurable model.