摘要：AbstractAn extension of the built-in self-scaling (BS) method to online estimation of impulse responses is proposed. It allows estimation to be carried out in real time, which is beneficial for online monitoring, in line with trends in Industry 4.0. The method achieves higher accuracy compared with the standard kernel-based (KB) estimation, and allows the estimate to be available before the end of the identification test. This leads to time savings which are important particularly for the industry. The relationship between the standard KB estimate and that of the BS method is expounded through a rotation matrix, which shows how the dc gain term in the BS technique adjusts the standard KB estimate. Results from a hot air blower experiment illustrate the main concepts introduced.
