Non-ideal step response identification modeling algorithm of the second-order delayed system

Ideal step signal is nonexistent in practical application and the input signals used in system identification are actually not ideal step signals. This introduced errors to the direct identification modeling method that based on the ideal step input assumption. Therefore, an improved algorithm based on non-ideal step signal input was proposed in this paper. A parameter represent the time of the non-ideal step signal was introduced in this algorithm, and it can be decided automatically during the model identification by using genetic optimization algorithm (GA), which means that this proposed identification modeling algorithm has higher precision and also can be used directly in the practical application. The effectiveness and applicability of this method has been demonstrated through the simulation model identification under the non-ideal step signal input.