Error Estimation of the Nonlinear Systems Simulation Using Convolution Integral

The present paper explores the error estimation for piecewise approximation of nonlinear systems using convolution integral. The base for the proposed method is the relationship between convolution integral and z-transform. The problem of error estimation is simplified using complete system decomposition to two elementary dynamic blocks - integral (zero pole) and first-order (one real pole). For each block, possible cases of changes in their parameters are considered- one parameter in the integration block and up to two parameters in the first-order block. Theoretical principles confirmed by computer simulation using the simple model also been presented.