Stefan Jakubek
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Stefan Jakubek has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Christoph Hametner
Institute of Mechanics and Mechatronics, TU Wien, Vienna, Austria
14 shared publications - Martin Kozek
Institute of Mechanics and Mechatronics, Vienna University of Technology, Vienna, Austria
5 shared publications - Alexander Schirrer
Institute of Mechanics and Mechatronics, TU Wien, Vienna, Austria
4 shared publications - Oliver König
AVL List GmbH, Graz, Austria
4 shared publications - Daniel Ritzberger
Institute of Mechanics and Mechatronics, TU Wien, Getreidemarkt 9/E325, 1060 Vienna, Austria
3 shared publications Martin Kozek
Institute of Mechanics and Mechatronics, TU Wien, Vienna, Austria
3 shared publications
Latest Publications
- Open Access
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 6686-6701
Two-Degree-of-Freedom Compound Split Transmission Control for a Helicopter Powertrain
This paper discusses the efficient real-time control of the transmission ratio for the main rotor of a compound split gearbox designed for helicopter powertrains. The gearbox consists of two electric machines with a battery power supply and a compound split system, which significantly enhances overall performance. A two-degree-of-freedom (2DOF) controller employs a robust, model-based linear approach to ensure real-time control for the compound split, utilizing a method to obtain optimal performance while adhering to system constraints. Additionally, an augmented linear quadratic regulator (LQR) is designed to accurately track reference speeds and to attenuate coupled powertrain/rotor vibrations. The design method allows the designer to specify a constraint matrix to effectively integrate various inputs and control strategies, including load management and energy optimization. Simulation results illustrate the efficiency of the proposed control strategy in various flight maneuvers. Additionally, the results substantiate the design of the compound split powertrain, thereby establishing a foundation for future advancements in electrified helicopter propulsion systems.
- 19 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 6686-6701
Open AccessTwo-Degree-of-Freedom Compound Split Transmission Control for a Helicopter Powertrain
This paper discusses the efficient real-time control of the transmission ratio for the main rotor of a compound split gearbox designed for helicopter powertrains. The gearbox consists of two electric machines with a battery power supply and a compound split system, which significantly enhances overall performance. A two-degree-of-freedom (2DOF) controller employs a robust, model-based linear approach to ensure real-time control for the compound split, utilizing a method to obtain optimal performance while adhering to system constraints. Additionally, an augmented linear quadratic regulator (LQR) is designed to accurately track reference speeds and to attenuate coupled powertrain/rotor vibrations. The design method allows the designer to specify a constraint matrix to effectively integrate various inputs and control strategies, including load management and energy optimization. Simulation results illustrate the efficiency of the proposed control strategy in various flight maneuvers. Additionally, the results substantiate the design of the compound split powertrain, thereby establishing a foundation for future advancements in electrified helicopter propulsion systems.
- Open Access
- 11 February 2026
- Published by Elsevier in SSRN Electronic Journal
On-Board Battery State of Health Estimation for Partial Charge Segments using Incremental Capacity Analysis
Accurately estimating the state of health of a battery is essential for the safe and efficient operation of electric vehicles. However, performing this estimation under real-world conditions poses significant challenges, as only limited measurement data - current, voltage, and temperature - are available during vehicle operation. This work presents a data-driven framework for state of health estimation based on automatic feature extraction. To reveal battery degradation patterns, incremental capacity curves derived from constant charge phases are analyzed. Through proper orthogonal decomposition, the dimensionality of the feature space is reduced while preserving essential information. These extracted features are then used to train a machine learning model for accurate state of health estimation. The framework can process partial charge segments from different state of charge regions, overcoming the limitation of requiring complete charge cycles. This capability makes the method particularly suitable for onboard implementation in electric vehicles. Validation is performed using data from large-scale aging experiments that include both structured test cycles and dynamic charging conditions. The model consistently captures degradation trends, confirming its robustness and suitability for use in onboard state of health monitoring of electric vehicles.
- 11 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessOn-Board Battery State of Health Estimation for Partial Charge Segments using Incremental Capacity Analysis
Accurately estimating the state of health of a battery is essential for the safe and efficient operation of electric vehicles. However, performing this estimation under real-world conditions poses significant challenges, as only limited measurement data - current, voltage, and temperature - are available during vehicle operation. This work presents a data-driven framework for state of health estimation based on automatic feature extraction. To reveal battery degradation patterns, incremental capacity curves derived from constant charge phases are analyzed. Through proper orthogonal decomposition, the dimensionality of the feature space is reduced while preserving essential information. These extracted features are then used to train a machine learning model for accurate state of health estimation. The framework can process partial charge segments from different state of charge regions, overcoming the limitation of requiring complete charge cycles. This capability makes the method particularly suitable for onboard implementation in electric vehicles. Validation is performed using data from large-scale aging experiments that include both structured test cycles and dynamic charging conditions. The model consistently captures degradation trends, confirming its robustness and suitability for use in onboard state of health monitoring of electric vehicles.
- Open Access
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 168344-168360
End-to-End Process Optimization in Polymer Extrusion Lines Using Model Predictive Control and Multi-Task Learning
Due to the inherent complexity of modern polymer extrusion lines caused by nonlinear, dynamic behavior with numerous influential factors, disturbances lead to higher scrap rates and downtimes. Unfortunately, conventional approaches to address this challenge rely on independent models and local control loops, neglecting the multistage characteristic of extrusion lines and therefore, cannot obtain a global optimum. Consequently, we propose a predictive control design based on a nonlinear and autoregressive multi-task learning model covering the entire extrusion line including local control loops and essential quality measurements. Training on an extended prediction horizon successfully addresses accumulating prediction errors and measurement noise. Due to its capability to predict state trajectories for up to 60 minutes, the proposed methodology enables effective model predictive control. Additionally, using an efficient method for adaptive error compensation based on previous error trajectories increases robustness substantially while the hierarchical control architecture supports global optimization and efficient local control loops. The proposed control design is validated on two polymer extrusion lines. The results show that changing operating points and optimizing process states, such as melt temperature, can be reliably achieved in spite of process disturbances. In comparison to baseline production periods, dominant oscillations are successfully damped by 62 %, reference values are closely followed and process variations are reduced by 41 % to 63 % improving product quality notably. A comparative analysis of disturbance rejection capabilities shows superior results while empirical analyses of stability and runtime further improve its closed-loop applicability. As a consequence, the proposed methodology for predictive control allows to steer extrusion processes accurately, address upcoming issues in advance and generally leads to more efficient processes.
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 168344-168360
Open AccessEnd-to-End Process Optimization in Polymer Extrusion Lines Using Model Predictive Control and Multi-Task Learning
Due to the inherent complexity of modern polymer extrusion lines caused by nonlinear, dynamic behavior with numerous influential factors, disturbances lead to higher scrap rates and downtimes. Unfortunately, conventional approaches to address this challenge rely on independent models and local control loops, neglecting the multistage characteristic of extrusion lines and therefore, cannot obtain a global optimum. Consequently, we propose a predictive control design based on a nonlinear and autoregressive multi-task learning model covering the entire extrusion line including local control loops and essential quality measurements. Training on an extended prediction horizon successfully addresses accumulating prediction errors and measurement noise. Due to its capability to predict state trajectories for up to 60 minutes, the proposed methodology enables effective model predictive control. Additionally, using an efficient method for adaptive error compensation based on previous error trajectories increases robustness substantially while the hierarchical control architecture supports global optimization and efficient local control loops. The proposed control design is validated on two polymer extrusion lines. The results show that changing operating points and optimizing process states, such as melt temperature, can be reliably achieved in spite of process disturbances. In comparison to baseline production periods, dominant oscillations are successfully damped by 62 %, reference values are closely followed and process variations are reduced by 41 % to 63 % improving product quality notably. A comparative analysis of disturbance rejection capabilities shows superior results while empirical analyses of stability and runtime further improve its closed-loop applicability. As a consequence, the proposed methodology for predictive control allows to steer extrusion processes accurately, address upcoming issues in advance and generally leads to more efficient processes.
- Open Access
- 19 May 2025
- Published by Springer Nature in Journal of Intelligent Manufacturing
- Vol. 37 (4) , 1701-1715
Multi-task learning with state propagation for quality forecasts in polymer extrusion lines
Modern polymer extrusion processes usually consist of several process steps, exhibit dynamic nonlinear behavior and depend on many external factors, for instance ambient conditions. Consequently, disturbances cannot be anticipated and extrusion lines are usually operated in a reactive manner leading to process fluctuations, quality deviations and increased material consumption. Unfortunately, existing approaches are either focused on single stage solutions which do not consider the process structure or rely on static models neglecting the dynamic behavior. Therefore, we propose a multi-task learning framework of individual (nonlinear) autoregressive state-models with exogenous inputs connected via a causality graph and jointly trained via physics-informed state propagation. Additionally, training the model on an extended forecasting horizon allows to capture and forecast the dynamic process behavior for a prolonged period. Moreover, the modular architecture can be adapted for any continuous production process with sequential and parallel material streams as well as feedback loops of existing controllers. The proposed method is validated with data from an actual pipe extrusion line and compared to end-to-end black-box models. The results show superior performance with accurate forecasts up to 120 min with a mean absolute percentage error of 0.09% for the pipe diameter representing a reduction of 13.9% to 22.3% compared to different baseline models. Furthermore, the effect of spurious correlations is mitigated through enforced physical dependencies improving the robustness of the resulting model. This allows to change the operating mode from reactive to proactive and to reduce the need for trial and error optimizations which leads to increased efficiency in complex extrusion lines.
- 19 May 2025
- Published by Springer Nature in Journal of Intelligent Manufacturing
- Vol. 37 (4) , 1701-1715
Open AccessMulti-task learning with state propagation for quality forecasts in polymer extrusion lines
Modern polymer extrusion processes usually consist of several process steps, exhibit dynamic nonlinear behavior and depend on many external factors, for instance ambient conditions. Consequently, disturbances cannot be anticipated and extrusion lines are usually operated in a reactive manner leading to process fluctuations, quality deviations and increased material consumption. Unfortunately, existing approaches are either focused on single stage solutions which do not consider the process structure or rely on static models neglecting the dynamic behavior. Therefore, we propose a multi-task learning framework of individual (nonlinear) autoregressive state-models with exogenous inputs connected via a causality graph and jointly trained via physics-informed state propagation. Additionally, training the model on an extended forecasting horizon allows to capture and forecast the dynamic process behavior for a prolonged period. Moreover, the modular architecture can be adapted for any continuous production process with sequential and parallel material streams as well as feedback loops of existing controllers. The proposed method is validated with data from an actual pipe extrusion line and compared to end-to-end black-box models. The results show superior performance with accurate forecasts up to 120 min with a mean absolute percentage error of 0.09% for the pipe diameter representing a reduction of 13.9% to 22.3% compared to different baseline models. Furthermore, the effect of spurious correlations is mitigated through enforced physical dependencies improving the robustness of the resulting model. This allows to change the operating mode from reactive to proactive and to reduce the need for trial and error optimizations which leads to increased efficiency in complex extrusion lines.
- Open Access
- 14 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 86968-86982
Experimental Evaluation of a 2DOF Controller for a Novel Multi-Temperature Transport System
Multi-temperature transport systems offer a significant opportunity to improve supply chain efficiency by enabling the simultaneous transport of goods with diverse temperature requirements. However, these systems present critical environmental challenges due to high energy consumption, refrigerant leakage, and the significant amount of goods being damaged or lost due to inadequate temperature control. To address these issues, this study proposes a novel indirect cascade multi-temperature transport system with two temperature-controlled compartments. The design features a compact, well-sealed refrigeration system with reduced refrigerant charge and utilizes the natural refrigerant propane. A nonlinear control strategy based on feedback linearization is introduced to decouple system dynamics and precisely regulate compartment temperatures. Specifically, a two-degree-of-freedom (2DOF) controller, consisting of a feedforward trajectory planning algorithm and a feedback controller, is implemented. Additionally, an optimization-based input transformation allows prioritization of specific compartments when actuator limitations prevent accurate tracking of all compartment temperatures. The control approach was experimentally evaluated on a test bed and compared with a conventional proportional-integral (PI) controller. Across the conducted experiments, the proposed 2DOF control law reduced temperature deviations during normal operation by 40% to 77%, clearly highlighting its quantitative superiority over the conventional approach.
- 14 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 86968-86982
Open AccessExperimental Evaluation of a 2DOF Controller for a Novel Multi-Temperature Transport System
Multi-temperature transport systems offer a significant opportunity to improve supply chain efficiency by enabling the simultaneous transport of goods with diverse temperature requirements. However, these systems present critical environmental challenges due to high energy consumption, refrigerant leakage, and the significant amount of goods being damaged or lost due to inadequate temperature control. To address these issues, this study proposes a novel indirect cascade multi-temperature transport system with two temperature-controlled compartments. The design features a compact, well-sealed refrigeration system with reduced refrigerant charge and utilizes the natural refrigerant propane. A nonlinear control strategy based on feedback linearization is introduced to decouple system dynamics and precisely regulate compartment temperatures. Specifically, a two-degree-of-freedom (2DOF) controller, consisting of a feedforward trajectory planning algorithm and a feedback controller, is implemented. Additionally, an optimization-based input transformation allows prioritization of specific compartments when actuator limitations prevent accurate tracking of all compartment temperatures. The control approach was experimentally evaluated on a test bed and compared with a conventional proportional-integral (PI) controller. Across the conducted experiments, the proposed 2DOF control law reduced temperature deviations during normal operation by 40% to 77%, clearly highlighting its quantitative superiority over the conventional approach.
- 11 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 72 (10) , 3000-3006
In-Vitro Hemocompatibility Evaluation of the HeartMate 3 Under Realistic Operating Conditions
Objective: Operating conditions significantly influence hemolysis generation in rotodynamic blood pumps (RBPs). Previous experiments conducted under constant operating conditions have demonstrated that lower flow rates are associated with a higher normalized index of hemolysis (NIH). However, in clinical scenarios, the pump flow rate fluctuates in response to the residual cardiac function. This study aims to investigate the effects of various pulsatile operating conditions on hemolysis generation and von Willebrand Factor (vWF) degradation of RBP. Methods: To investigate these conditions, pulsatile flow experiments were conducted for 12 hours using HeartMate 3 (HM3, Abbott Inc, USA) in a hybrid mock circulatory loop with citrated human blood from hemochromatosis patients. Three operating conditions (high, low, and no residual cardiac function) were examined under two pump speed settings: normal (5400 rpm, 4.3 L/min) and low (4800 rpm, 2.5 L/min). Hemolysis was assessed by measuring delta free hemoglobin every 30 minutes (dfHb30 min) and calculating NIH. High molecular weight (HMW) vWF multimer degradation was assessed using immunoblotting. Results: There was no significant difference in dfHb30 min (p>0.388), NIH (p > 0.382), and HMW vWF multimers degradation (p > 0.364) between the three operating conditions, but differences in these parameters were observed between the normal and low speed settings. Meanwhile, a consistent trend in the hemolysis outcomes was observed with slightly elevated hemolysis in no and high pulsatility conditions of both speed settings. Conclusion: Hemocompatibility of the HM3 is not significantly affected by periodic high/low flow or backflows through the pump in in-vitro evaluation but rather by the pump operating condition: flow rate and pump speed.
- 11 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 72 (10) , 3000-3006
In-Vitro Hemocompatibility Evaluation of the HeartMate 3 Under Realistic Operating Conditions
Objective: Operating conditions significantly influence hemolysis generation in rotodynamic blood pumps (RBPs). Previous experiments conducted under constant operating conditions have demonstrated that lower flow rates are associated with a higher normalized index of hemolysis (NIH). However, in clinical scenarios, the pump flow rate fluctuates in response to the residual cardiac function. This study aims to investigate the effects of various pulsatile operating conditions on hemolysis generation and von Willebrand Factor (vWF) degradation of RBP. Methods: To investigate these conditions, pulsatile flow experiments were conducted for 12 hours using HeartMate 3 (HM3, Abbott Inc, USA) in a hybrid mock circulatory loop with citrated human blood from hemochromatosis patients. Three operating conditions (high, low, and no residual cardiac function) were examined under two pump speed settings: normal (5400 rpm, 4.3 L/min) and low (4800 rpm, 2.5 L/min). Hemolysis was assessed by measuring delta free hemoglobin every 30 minutes (dfHb30 min) and calculating NIH. High molecular weight (HMW) vWF multimer degradation was assessed using immunoblotting. Results: There was no significant difference in dfHb30 min (p>0.388), NIH (p > 0.382), and HMW vWF multimers degradation (p > 0.364) between the three operating conditions, but differences in these parameters were observed between the normal and low speed settings. Meanwhile, a consistent trend in the hemolysis outcomes was observed with slightly elevated hemolysis in no and high pulsatility conditions of both speed settings. Conclusion: Hemocompatibility of the HM3 is not significantly affected by periodic high/low flow or backflows through the pump in in-vitro evaluation but rather by the pump operating condition: flow rate and pump speed.
- Open Access
- 2 April 2025
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 113 (14) , 18265-18289
Nonlinear distributed-parameter observer design for efficient estimation of internal temperature profiles in polymer electrolyte membrane fuel cells
This paper presents a methodology to design a nonlinear distributed-parameter observer to estimate internal temperature profiles in polymer electrolyte membrane fuel cells. Accurate knowledge of the spatial temperature distributions allows beneficial insight into the cell’s condition since the temperature is strongly coupled to other cell states. The extended Kalman filter-based observer employs a high-fidelity nonlinear non-isothermal distributed-parameter model for predicting internal temperature distributions. A reduced-order model, preserving the major dynamics and coupling effects of the original model, is utilized for real-time capable and feasible state correction. To improve the estimation, the observer scheme is augmented to simultaneously estimate selected material parameters along the temperature profiles online. In a simulation experiment, the observer is compared to an open-loop simulation benchmark. Further, the consistency of the estimation is investigated. The observer methodology is validated using a high-resolution simulated reality since internal temperature distributions are inaccessible in a real system. The validation demonstrates the observer’s accuracy in estimating the unmeasurable internal temperature distributions. The results substantiate the observer’s advantage over open-loop model simulations and outline the importance of temperature profile estimation for meaningful condition monitoring.
- 2 April 2025
- Published by Springer Nature in Nonlinear Dynamics
- Vol. 113 (14) , 18265-18289
Open AccessNonlinear distributed-parameter observer design for efficient estimation of internal temperature profiles in polymer electrolyte membrane fuel cells
This paper presents a methodology to design a nonlinear distributed-parameter observer to estimate internal temperature profiles in polymer electrolyte membrane fuel cells. Accurate knowledge of the spatial temperature distributions allows beneficial insight into the cell’s condition since the temperature is strongly coupled to other cell states. The extended Kalman filter-based observer employs a high-fidelity nonlinear non-isothermal distributed-parameter model for predicting internal temperature distributions. A reduced-order model, preserving the major dynamics and coupling effects of the original model, is utilized for real-time capable and feasible state correction. To improve the estimation, the observer scheme is augmented to simultaneously estimate selected material parameters along the temperature profiles online. In a simulation experiment, the observer is compared to an open-loop simulation benchmark. Further, the consistency of the estimation is investigated. The observer methodology is validated using a high-resolution simulated reality since internal temperature distributions are inaccessible in a real system. The validation demonstrates the observer’s accuracy in estimating the unmeasurable internal temperature distributions. The results substantiate the observer’s advantage over open-loop model simulations and outline the importance of temperature profile estimation for meaningful condition monitoring.
- Open Access
- 7 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 10 (8) , 4153-4170
Agile Mixed-Integer-Based Lane-Change MPC for Collision-Free and Efficient Autonomous Driving
Obstacle avoidance and lane-change functionalities are crucial requirements for automated driving that are usually developed to excel in specific road scenarios and therefore, to some extent, lack agility and versatility. This work proposes a versatile vehicle motion controller for connected and autonomous vehicles (CAVs) that, independently of the encountered traffic scenario, realizes safe and efficient lane-change and overtaking maneuvers while implicitly guaranteeing precise obstacle avoidance at all times. This is achieved by introducing a two-layer model predictive control architecture that utilizes mixed-integer-based obstacle avoidance and lane selection formulations. The linear time-invariant optimal control problems are efficiently formulated in Frenet coordinates, incorporate the position predictions of surrounding traffic participants, and guarantee globally optimal solutions. The proposed architecture combines two structurally distinct policies, namely velocity and time-gap tracking, to achieve efficient and safe maneuvering. A method to reduce the computational complexity of the implemented mixed-integer quadratic programming (MIQP) problems is developed. The controller's agility and robustness with respect to multiple complex traffic scenarios are tested in detailed traffic simulations, as well as validated via high-fidelity co-simulations with the CARLA Simulator.
- 7 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 10 (8) , 4153-4170
Open AccessAgile Mixed-Integer-Based Lane-Change MPC for Collision-Free and Efficient Autonomous Driving
Obstacle avoidance and lane-change functionalities are crucial requirements for automated driving that are usually developed to excel in specific road scenarios and therefore, to some extent, lack agility and versatility. This work proposes a versatile vehicle motion controller for connected and autonomous vehicles (CAVs) that, independently of the encountered traffic scenario, realizes safe and efficient lane-change and overtaking maneuvers while implicitly guaranteeing precise obstacle avoidance at all times. This is achieved by introducing a two-layer model predictive control architecture that utilizes mixed-integer-based obstacle avoidance and lane selection formulations. The linear time-invariant optimal control problems are efficiently formulated in Frenet coordinates, incorporate the position predictions of surrounding traffic participants, and guarantee globally optimal solutions. The proposed architecture combines two structurally distinct policies, namely velocity and time-gap tracking, to achieve efficient and safe maneuvering. A method to reduce the computational complexity of the implemented mixed-integer quadratic programming (MIQP) problems is developed. The controller's agility and robustness with respect to multiple complex traffic scenarios are tested in detailed traffic simulations, as well as validated via high-fidelity co-simulations with the CARLA Simulator.
- Open Access
- 8 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (11) , 16364-16374
Predictive Energy Management Strategy for Fuel Cell Vehicles Combining Long-Term and Short-Term Forecasts
Fuel cell electric vehicles are usually hybrid vehicles requiring an energy management strategy (EMS) to determine the power split between the fuel cell system and a battery. The performance of an EMS can be improved by taking into account forecasts of the vehicle velocity. Simple estimates derived from static route information, e.g., speed limits, can already provide a significant performance increase because of being available before departure and for the entire driving mission. However, such long-term predictions can deviate considerably from the actual velocity because of dynamic influences, such as traffic, roadworks, or weather. Here, short-term predictions from vehicular communication systems provide more accurate real-time information and allow the EMS to react better to the actual driving conditions. This article proposes a predictive EMS that efficiently combines the information of long-term and short-term forecasts. Before departure, a dynamic programming algorithm optimizes the energy management based on a-priori available route data yielding a distance-based map describing the optimal cost-to-go. While driving, a model predictive controller (MPC) optimizes the energy management online considering the information of the short-term prediction and including the optimal cost-to-go as terminal cost. A computationally efficient linear MPC implementation is proposed, and the significant performance benefit over an MPC that tracks an optimized battery state-of-charge reference is demonstrated in a numerical study.
- 8 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (11) , 16364-16374
Open AccessPredictive Energy Management Strategy for Fuel Cell Vehicles Combining Long-Term and Short-Term Forecasts
Fuel cell electric vehicles are usually hybrid vehicles requiring an energy management strategy (EMS) to determine the power split between the fuel cell system and a battery. The performance of an EMS can be improved by taking into account forecasts of the vehicle velocity. Simple estimates derived from static route information, e.g., speed limits, can already provide a significant performance increase because of being available before departure and for the entire driving mission. However, such long-term predictions can deviate considerably from the actual velocity because of dynamic influences, such as traffic, roadworks, or weather. Here, short-term predictions from vehicular communication systems provide more accurate real-time information and allow the EMS to react better to the actual driving conditions. This article proposes a predictive EMS that efficiently combines the information of long-term and short-term forecasts. Before departure, a dynamic programming algorithm optimizes the energy management based on a-priori available route data yielding a distance-based map describing the optimal cost-to-go. While driving, a model predictive controller (MPC) optimizes the energy management online considering the information of the short-term prediction and including the optimal cost-to-go as terminal cost. A computationally efficient linear MPC implementation is proposed, and the significant performance benefit over an MPC that tracks an optimized battery state-of-charge reference is demonstrated in a numerical study.
- Open Access
- 29 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (2) , 841-855
Agent-Based Decentralized Model Predictive Control for Plants With Multiple Identical Actuators
This article proposes a decentralized model predictive control (DMPC) algorithm without communication for systems consisting of multiple identical, independent actuators acting on a single central plant. The particular system design is relevant for applications where modularity is paramount and for highly dynamic systems, e.g., battery emulator systems controlled by multiple dc–dc converter modules. Each agent, consisting of an actuator and a DMPC, controls a virtually scaled version of the plant to implicitly consider the effects of other agents. The set of DMPCs achieves the same plant performance as a corresponding centralized model predictive controller (CMPC) in unconstrained operation. Also, the states of the independent agents converge toward the globally optimal CMPC solution. This is obtained by dividing the CMPC’s objective function into local objective functions related to the subsystems. The optimality and stability of the DMPC in unconstrained operation are shown analytically. The stability of control input-constrained operation is analyzed by computing the region of attraction. Numerical studies of a battery emulator system compare the performance of the DMPC with the global optimum in detail.
- 29 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Control Systems Technology
- Vol. 31 (2) , 841-855
Open AccessAgent-Based Decentralized Model Predictive Control for Plants With Multiple Identical Actuators
This article proposes a decentralized model predictive control (DMPC) algorithm without communication for systems consisting of multiple identical, independent actuators acting on a single central plant. The particular system design is relevant for applications where modularity is paramount and for highly dynamic systems, e.g., battery emulator systems controlled by multiple dc–dc converter modules. Each agent, consisting of an actuator and a DMPC, controls a virtually scaled version of the plant to implicitly consider the effects of other agents. The set of DMPCs achieves the same plant performance as a corresponding centralized model predictive controller (CMPC) in unconstrained operation. Also, the states of the independent agents converge toward the globally optimal CMPC solution. This is obtained by dividing the CMPC’s objective function into local objective functions related to the subsystems. The optimality and stability of the DMPC in unconstrained operation are shown analytically. The stability of control input-constrained operation is analyzed by computing the region of attraction. Numerical studies of a battery emulator system compare the performance of the DMPC with the global optimum in detail.