การประยุกต์ใช้สัญญาณสุ่มขาวแบบเกาส์สำหรับการสร้างแบบจำลองระบบไม่เชิงเส้น

Abstract

This research presents the separation of the linear transfer function of nonlinear systems based on the perpendicular properties of the Gaussian White Noise signal as the input signal. The input and output signals are continuously collected in real time to average the signal and to avoid excessive memory usage, Circular Buffer has been applied.  The Goyder’s technique performed well in separation linear signal for the mathematical modeling test. The transfer function also produced accurate results with a 0.003% error when it was used to calculate the linear system parameters.  This method's drawback is that the signal must be averaged for a considerable amount of time and the transfer function's quality must be visually examined. When the transfer function becomes steady, the signal measurement will come to an end.