Jakub Dokoupil

Jakub Dokoupil

Auto-generated profile, based on publicly available information

Jakub Dokoupil

Brno University of Technology

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 September 2016 in International Journal of Control
Verified
Journal Article
Closed access
ŁRŁukasz Ryba
Łukasz Ryba
JDJakub Dokoupil
Jakub Dokoupil
AVAlina Voda
Alina Voda
GBGildas BESANCON
Gildas BESANCON
This paper presents a novel adaptive approach for hysteresis compensation applied to a piezoelectric actuator in one axis of an STM-like lab-made micro-/nanopositioning platform. The idea is to identify a compensating static parametric model, which imitates directly the inverse model of the nonlinearity. In this way, the approach is less complex than those based on model inversion. In addition, the identification is made online, allowing to consider a simple polynomial model, and to adapt its parameters according to the actual hysteresis curve which is faced (ascending or descending path, varying input amplitude, etc.). In order to be able to track possibly fast parameter variations, an original adaptation algorithm is proposed within the Bayesian framework, and including an exponential forgetting factor with optimal data-driven tuning. Illustrative experimental results are finally presented for tracking both triangular and sinusoidal reference signals with varying amplitude.

5 July 2022 in IEEE Transactions on Automatic Control
Verified
Journal Article
Open access
JDJakub Dokoupil
Jakub Dokoupil
PVPavel Václavek
Pavel Václavek
The real-time estimation of the time-varying Hammerstein system by using a noniterative learning schema is considered and extended to incorporate a matrix forgetting factor. The estimation is cast in a variational-Bayes framework to best emulate the original posterior distribution of the parameters within the set of distributions with feasible moments. The recursive concept we propose approximates the exact posterior comprising undistorted information about the estimated parameters. In many practical settings, the incomplete model of parameter variations is compensated by forgetting of obsolete information. As a rule, the forgetting operation is initiated by the inclusion of an appropriate prediction alternative into the time update. It is shown that the careful formulation of the prediction alternative, which relies on Bayesian conditioning, results in partial forgetting. This article inspects two options with respect to the order of the conditioning in the posterior, which proves vital in the successful localization of the source of inconsistency in the data-generating process. The geometric mean of the discussed alternatives then modifies recursive learning through the matrix forgetting factor. We adopt the decision-making approach to revisit the posterior uncertainty by dynamically allocating the probability to each of the prediction alternatives to be combined.

20 January 2017 in IEEE Transactions on Automatic Control
Verified
Journal Article
Closed access
JDJakub Dokoupil
Jakub Dokoupil
AVAlina Voda
Alina Voda
PVPavel Vaclavek
Pavel Vaclavek
This technical note concerns the problem of variable regularized estimation of time-varying nonlinear systems from the Bayesian viewpoint. The questions of how to impose the posterior information being variably regularized and how to forget this information are carefully discussed. The estimator design adopts the strategy of the iterated Kalman filter but differs in that, instead of the separated moments of the linearized system, only the augmented covariance matrix is updated. To suppress obsolete information, a decision problem involving the Kullback–Leibler divergence is solved. The decision provides the best combination of a pair of time-evolution model hypotheses in terms of the geometric mean. As a result, exponential forgetting with the adaptively tuned factor is inserted into the estimation process. The regularization of the investigated statistics is induced through the processing of externally supplied information. The presented estimator allows for absolute discarding or, conversely, retention of external information produced in terms of the Normal–Wishart distribution.

1 December 2015
Verified
Conference Paper
Closed access
JDJakub Dokoupil
Jakub Dokoupil
MPMilan Papez
Milan Papez
PVPavel Vaclavek
Pavel Vaclavek
A novel growing-window recursive procedure for Kalman filter comparison is proposed based on the Bayesian inference principle. This procedure is capable of processing unlimited growth of the uncertainty of the initial parameter settings, which is a characteristic of Kalman type algorithms. The present paper applies the suggested procedure to assess the degree of support for the state point estimates generated by Kalman filters differing in their system model descriptions. The algebraic form of the comparison algorithm covers the situation when the covariance of the measurement noise is known as well as is unknown and the normalized covariance matrix of the process noise is always available. In this respect, the Kalman filter is formulated here as recursive learning of the sufficient statistics of the Normal and Normal-inverse-Wishart distributions.

1 December 2019
Verified
Conference Paper
Closed access
JDJakub Dokoupil
Jakub Dokoupil
PVPavel Vaclavek
Pavel Vaclavek
The paper addresses the problem of filtering the state of a normal dynamical process with a dependency between the process and the measurement noise variables in the presence of an inaccurate model description. As regards the time occurrence of the noise dependency, we discuss the dependency structure where both the variables are correlated at the same time. An adaptive formulation of the Kalman filter (KF) is designed in order to mitigate the impact of the process model uncertainty on the degradation of the filter performance. The filter we propose exploits the collaborative decision to introduce a variable forgetting factor into the time update to reduce artificially the effect of obsolete knowledge on the filtering solution. Within the decision-making rules, a loss functional quantifying the time update is constructed to optimally combine the prediction alternatives possessing the form of the normal probability density function (pdf). The result is an adjustment of the Kalman gain matrix in response to empirically confirmed performance.

1 June 2016
Verified
Conference Paper
Closed access
JDJakub Dokoupil
Jakub Dokoupil
PVPavel Vaclavek
Pavel Vaclavek
This paper addresses the adaptive estimation problem of time-varying systems in the Bayesian framework. The version of exponential forgetting with the variable factor is derived by solving the decision problem where the Kullback-Leibler divergence is used. This divergence is applied to evaluate the distance of two antagonistic model hypotheses from the model of parameter variations. The first hypothesis assumes no parameter changes, while the second one admits that the parameters may arbitrarily evolve throughout the parameter space. In this respect, the forgetting factor is interpreted as the probability that the first hypothesis meets the reality. This concept brings another technique into the class of self-tuned forgetting strategies for the discarding of obsolete information. The developed concept of forgetting is designed to complement the data learning process propagating the statistics of the Normal-Wishart distribution.