Bernardo Barancelli Schwedersky
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Bernardo Barancelli Schwedersky has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Rodolfo Cesar Costa Flesch
Department of Automation and Systems Engineering, Universidade Federal de Santa Catarina, Florianópolis 88040-900, SC, Brazil
6 shared publications - João Paulo Zomer Machado
Department of Automation and Systems Engineering, Federal University of Santa Catarina, Florianópolis, Brazil
2 shared publications - Ahryman Seixas Busse de Siqueira Nascimento
Department of Automation and Systems Engineering, Universidade Federal de Santa Catarina, Campus Universitário Reitor João David Ferreira Lima, Florianópolis, 88040-900, SC, Brazil
1 shared publication - Armando Albertazzi
Universidade Federal de Santa Catarina
1 shared publication - Bernardo C. F. de Oliveira
Campus Universitário UFSC, Universidade Federal De Santa Catarina, Florianópolis, Brazil
1 shared publication - Cassiano Montibeller
Programa de Ps-Graduao em Automao e Sistemas, Universidade Federal de Santa Catarina, Florianpolis, Santa Catarina, Brazil
1 shared publication
Latest Publications
- Open Access
- 11 February 2026
- Published by Elsevier in ISA Transactions
- Vol. 174, 287-297
Soft-sensing for compressor test time reduction with time-delay neural networks
- 11 February 2026
- Published by Elsevier in ISA Transactions
- Vol. 174, 287-297
Open AccessSoft-sensing for compressor test time reduction with time-delay neural networks
- Open Access
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 146745-146756
Practical Nonlinear Model Predictive Control With Kolmogorov–Arnold Network Models
This paper presents a novel approach to practical nonlinear model predictive control (PNMPC) using Kolmogorov–Arnold networks (KANs) as prediction models. KANs are based on the Kolmogorov–Arnold superposition theorem, enabling efficient nonlinear system modeling with fewer parameters compared to traditional neural networks. This characteristic makes KANs particularly suitable for nonlinear model predictive control (NMPC), where computational efficiency is critical. This paper presents a method to efficiently identify nonlinear models from data, proposing a time-delay KAN (TDKAN) architecture, which consists of a KAN associated with tap delays in its input. Furthermore, it proposes the integration of this novel model architecture into a PNMPC algorithm, with the aim of improving control accuracy while reducing computational burden. The method is validated using a pH neutralization process, a benchmark for nonlinear control problems. The results show that the TDKAN-based PNMPC outperforms other PNMPC approaches, such as those that use echo-state networks and long short-term memory networks, providing better setpoint tracking, lower mean squared error, and faster computation times, which demonstrate the potential of KANs for prediction in practical NMPC applications.
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 146745-146756
Open AccessPractical Nonlinear Model Predictive Control With Kolmogorov–Arnold Network Models
This paper presents a novel approach to practical nonlinear model predictive control (PNMPC) using Kolmogorov–Arnold networks (KANs) as prediction models. KANs are based on the Kolmogorov–Arnold superposition theorem, enabling efficient nonlinear system modeling with fewer parameters compared to traditional neural networks. This characteristic makes KANs particularly suitable for nonlinear model predictive control (NMPC), where computational efficiency is critical. This paper presents a method to efficiently identify nonlinear models from data, proposing a time-delay KAN (TDKAN) architecture, which consists of a KAN associated with tap delays in its input. Furthermore, it proposes the integration of this novel model architecture into a PNMPC algorithm, with the aim of improving control accuracy while reducing computational burden. The method is validated using a pH neutralization process, a benchmark for nonlinear control problems. The results show that the TDKAN-based PNMPC outperforms other PNMPC approaches, such as those that use echo-state networks and long short-term memory networks, providing better setpoint tracking, lower mean squared error, and faster computation times, which demonstrate the potential of KANs for prediction in practical NMPC applications.
- Open Access
- 25 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 77879-77892
Intelligent Air Gap Segmentation in Electric Motor Color Images Using a Self-Organizing Map
Image segmentation plays an important role in automating various inspection procedures that are still performed manually. In the production of electric motors, a relevant inspection task involves examining the air gap between the rotor and the stator to identify common defects. This paper introduces an intelligent approach based on a self-organizing map (SOM) that is trained individually for each image to automatically segment the air gap region in electric motor images. This procedure eliminates the need for a large dataset of manually segmented images, which is typically required by most of the intelligent segmentation methods in the literature. The procedure begins with a polar transformation of the motor image, which helps to identify the region of interest for segmentation. The selected region is then used to build a training dataset for the SOM. Each training example consists of a frame extracted around every pixel in the region. Once map training is complete, an automated procedure extracts the air gap segment from the map output by selecting the pixel cluster that corresponds to the air gap. Three configurations of this approach were evaluated, exploring the impact of using different color spaces (RGB, LAB, and HSV). To evaluate the proposed approach, six motor samples were tested, using different illumination conditions. The results were compared with two unsupervised segmentation methods: Otsu’s method and K-Means. The proposed approach outperformed both baseline methods in terms of evaluated metrics, achieving a 65% higher intersection-over-union score than the best baseline method (K-Means). This study demonstrates the efficacy of the proposed image segmentation method, particularly in scenarios where only a limited number of images are available, making it impractical to train a supervised model.
- 25 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 77879-77892
Open AccessIntelligent Air Gap Segmentation in Electric Motor Color Images Using a Self-Organizing Map
Image segmentation plays an important role in automating various inspection procedures that are still performed manually. In the production of electric motors, a relevant inspection task involves examining the air gap between the rotor and the stator to identify common defects. This paper introduces an intelligent approach based on a self-organizing map (SOM) that is trained individually for each image to automatically segment the air gap region in electric motor images. This procedure eliminates the need for a large dataset of manually segmented images, which is typically required by most of the intelligent segmentation methods in the literature. The procedure begins with a polar transformation of the motor image, which helps to identify the region of interest for segmentation. The selected region is then used to build a training dataset for the SOM. Each training example consists of a frame extracted around every pixel in the region. Once map training is complete, an automated procedure extracts the air gap segment from the map output by selecting the pixel cluster that corresponds to the air gap. Three configurations of this approach were evaluated, exploring the impact of using different color spaces (RGB, LAB, and HSV). To evaluate the proposed approach, six motor samples were tested, using different illumination conditions. The results were compared with two unsupervised segmentation methods: Otsu’s method and K-Means. The proposed approach outperformed both baseline methods in terms of evaluated metrics, achieving a 65% higher intersection-over-union score than the best baseline method (K-Means). This study demonstrates the efficacy of the proposed image segmentation method, particularly in scenarios where only a limited number of images are available, making it impractical to train a supervised model.
- 1 February 2022
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 108
Echo state networks for online, multi-step MPC relevant identification
- 1 February 2022
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 108
Echo state networks for online, multi-step MPC relevant identification
- 3 January 2022
- Published by Springer Nature in Journal of Control, Automation and Electrical Systems
- Vol. 33 (3) , 743-754
Nonlinear MIMO System Identification with Echo-State Networks
- 3 January 2022
- Published by Springer Nature in Journal of Control, Automation and Electrical Systems
- Vol. 33 (3) , 743-754
Nonlinear MIMO System Identification with Echo-State Networks
- 8 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 33 (6) , 2605-2614
Adaptive Practical Nonlinear Model Predictive Control for Echo State Network Models
This article proposes an adaptive practical nonlinear model predictive (NMPC) control algorithm which uses an echo state network (ESN) estimated online as a process model. In the proposed control algorithm, the ESN readout parameters are estimated online using a recursive least-squares method that considers an adaptive directional forgetting factor. The ESN model is used to obtain online a nonlinear prediction of the system free response, and a linearized version of the neural model is obtained at each sampling time to get a local approximation of the system step response, which is used to build the dynamic matrix of the system. The proposed controller was evaluated in a benchmark conical tank level control problem, and the results were compared with three baseline controllers. The proposed approach achieved similar results as the ones obtained by its nonadaptive baseline version in a scenario with the process operating with the nominal parameters, and outperformed all baseline algorithms in a scenario with process parameter changes. Additionally, the computational time required by the proposed algorithm was one-tenth of that required by the baseline NMPC, which shows that the proposed algorithm is suitable to implement state-of-the-art adaptive NMPC in a computationally affordable manner.
- 8 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 33 (6) , 2605-2614
Adaptive Practical Nonlinear Model Predictive Control for Echo State Network Models
This article proposes an adaptive practical nonlinear model predictive (NMPC) control algorithm which uses an echo state network (ESN) estimated online as a process model. In the proposed control algorithm, the ESN readout parameters are estimated online using a recursive least-squares method that considers an adaptive directional forgetting factor. The ESN model is used to obtain online a nonlinear prediction of the system free response, and a linearized version of the neural model is obtained at each sampling time to get a local approximation of the system step response, which is used to build the dynamic matrix of the system. The proposed controller was evaluated in a benchmark conical tank level control problem, and the results were compared with three baseline controllers. The proposed approach achieved similar results as the ones obtained by its nonadaptive baseline version in a scenario with the process operating with the nominal parameters, and outperformed all baseline algorithms in a scenario with process parameter changes. Additionally, the computational time required by the proposed algorithm was one-tenth of that required by the baseline NMPC, which shows that the proposed algorithm is suitable to implement state-of-the-art adaptive NMPC in a computationally affordable manner.