Differentiator application in altitude control for an indoor blimp robot

Abstract

This paper presents design of altitude controller with disturbance compensation for an indoor blimp robot and its realisation. Due to hardware restrictions, the altitude control behaviour of blimp is modelled as a switched system with a constant time-delay complemented with uncertain bounded disturbances. In order to achieve state estimation, four differentiators are applied and compared, then HOMD (homogeneous finite-time) differentiator is chosen as an observer for vertical velocity and switching signal estimation. Next, a predictor-based controller is conceived, and in order to compensate the perturbation, the method for disturbance evaluation is designed still with the help of HOMD differentiator. Control scheme is implemented by Matlab Simulink, and finally, the performance of blimp altitude controller is verified in experiments.

Comments on ‘Differentiator application in altitude control for an indoor blimp robot’

“…Here we set T = 0.5s in (a) and T = 2.5s in (b). Comparison with the corresponding figures 2 and 3 of Wang et al (2018) shows that their conclusions need to be reconsidered in the light of the comments above. …”

“…The simulations in Wang et al (2018) are reproduced below using these causal and non delayed estimators (see figures i. and ii. where HOMD, HOSM and HG correspond to homogeneous finite-time, high-order sliding mode and high-gain differentiators, respectively).…”

“…The algebraic estimations in equation (12) are derived from Mboup, Join, and Fliess (2007) which has been significantly improved and completed by Mboup et al (2009), and this last reference is not cited in Wang et al (2018). The consequence is an inappropriate application of the so-called ALIEN differentiators.…”

“…The algebraic differentiators admit both causal and anti-causal implementations. The estimators in equation (12) of Wang et al (2018) seem to implement the anti-causal versions.…”

“…In a recent work Wang et al (2018), the algebraic differentiators of Mboup, Join, and Fliess (2009) have been applied for the signal denoising and first-order derivative estimation of a signal x(t) from its noisy measurement y(t) = x(t) + noise. However, an improper application of the method resulted in unfair comparisons with other methods.…”