Efficient nonparametric identification for high-precision motion systems: A practical comparison based on a medical X-ray system

Non-parametric identification of multivariable systems: A local rational modeling approach with application to a vibration isolation benchmark

“…However, for clarity, such weighting factors have been omitted in the present paper. It has been shown, e.g, in [17,22,24], that the local polynomial method leads to an improved FRF estimate, especially in cases where there is a significant transient contribution. This is most relevant when the available measurement time is limited or when there are lightly damped resonances.…”

“…In [20], the LRM is formulated for single-input single-output (SISO) systems and uses a linear cost criterion, in the sense of [21], to estimate the local rational models. The use of these local rational models instead of polynomials has been shown to lead to significant improvements in the estimation quality for systems containing lightly damped resonances, [22][23][24]. For a recent overview of FRF identification methods applied to a lightly-damped mechanical positioning system, see, e.g., [24].…”

“…The use of these local rational models instead of polynomials has been shown to lead to significant improvements in the estimation quality for systems containing lightly damped resonances, [22][23][24]. For a recent overview of FRF identification methods applied to a lightly-damped mechanical positioning system, see, e.g., [24].…”

Best Linear Approximation of Nonlinear Continuous-Time Systems Subject to Process Noise and Operating in Feedback

“…Hence, it is important to analyze the impact of the process noise on the BLA. Beside control applications [5], [6] and amplifiers operating in closed loop [7], feedback is present in any experimental setup where the plant is excited by a non-ideal actuator [8]. It emphasizes the importance of handling nonlinear systems subject to process noise and operating in a closed loop [see Figure 1].…”

“…Using the BLA one can easily check the validity of the linear framework in practical applications such as, for example, operational amplifiers [7], industrial robots [5], bit-errorrate measurements in telecommunication [9], characterization of lithium ion batteries [10], control of a medical X-ray system [6], voltage instrument transformers [11], and current transformers [12]. In addition, the dependence of the BLA on the excitation power spectrum also provides some guidance for nonlinear model selection [13], [14].…”