Sai Tej Paruchuri
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
Research Subjects
Sai Tej Paruchuri has a total of 65 co-authors. Here are the top co-authors by number of shared publications:
- Andrew J. Kurdila
Department of Mechanical Engineering, Virginia Tech, Blacksburg, VA, USA
10 shared publications - Tariq Rafiq
Lehigh University
9 shared publications Eugenio Schuster
Lehigh University
7 shared publications- Eugenio Schuster
Department of Mechanical Engineering and Mechanics, Lehigh University, Bethlehem, PA, USA
7 shared publications - Zibo Wang
Department of Mechanical Engineering and Mechanics, Lehigh University, Bethlehem, PA 18015, USA
6 shared publications Andrew J. Kurdila
Virginia Tech: Virginia Polytechnic Institute and State University
5 shared publications
Latest Publications
- Open Access
- 31 July 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68  (8) , 085001
Experimental demonstration of reactor-relevant adaptive density control on the DIII-D tokamak with coordinated gas and pellet fueling
Precise regulation of plasma density is crucial for achieving and maintaining fusion-relevant conditions in reactor-grade tokamaks. Tokamak reactors will utilize both gas puffing and pellet injection as standard fueling mechanisms. However, the coordinated use of gas puffing and pellet injection within feedback control frameworks is largely unexplored, underscoring the necessity to develop and validate dedicated strategies on existing machines. These strategies must address the various challenges associated with both actuators, including actuation delays, unknown fueling efficiencies, and the coexistence of continuous-time and discrete-time dynamics. To overcome these challenges, which surpass the capabilities of traditional empirically tuned proportional-integral-derivative (PID) control, an indirect adaptive control algorithm is proposed in this work for the regulation of the line-averaged electron density through coordinated gas puffing and pellet injection. After initial validation in simulations with a multi-reservoir global particle model, the controller was successfully implemented and tested on the DIII-D tokamak. Experimental results demonstrate robust density tracking under reactor-relevant scenarios, showcasing the controller’s ability to seamlessly coordinate actuators while handling disturbances and evolving actuator constraints. This work provides a critical step forward in the control of future fusion reactors.
- 31 July 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68  (8) , 085001
Open AccessExperimental demonstration of reactor-relevant adaptive density control on the DIII-D tokamak with coordinated gas and pellet fueling
Precise regulation of plasma density is crucial for achieving and maintaining fusion-relevant conditions in reactor-grade tokamaks. Tokamak reactors will utilize both gas puffing and pellet injection as standard fueling mechanisms. However, the coordinated use of gas puffing and pellet injection within feedback control frameworks is largely unexplored, underscoring the necessity to develop and validate dedicated strategies on existing machines. These strategies must address the various challenges associated with both actuators, including actuation delays, unknown fueling efficiencies, and the coexistence of continuous-time and discrete-time dynamics. To overcome these challenges, which surpass the capabilities of traditional empirically tuned proportional-integral-derivative (PID) control, an indirect adaptive control algorithm is proposed in this work for the regulation of the line-averaged electron density through coordinated gas puffing and pellet injection. After initial validation in simulations with a multi-reservoir global particle model, the controller was successfully implemented and tested on the DIII-D tokamak. Experimental results demonstrate robust density tracking under reactor-relevant scenarios, showcasing the controller’s ability to seamlessly coordinate actuators while handling disturbances and evolving actuator constraints. This work provides a critical step forward in the control of future fusion reactors.
- Open Access
- 27 October 2025
- Published by Springer Nature in Research Square
Discrete-Time Dynamical Systems Characterization via Invariance and Approximation of Koopman Operators and Operator-Valued Kernels
This paper presents approaches based on native space theory and Koopman operators to characterize the dynamics of nonlinear, discrete-time dynamical models employing measured data only. Given approximation schemes of the plant dynamics based on Koopman operators contained in vector-valued reproducing kernel Hilbert spaces (vRKHSs), we deduce rates of convergence for these schemes. In particular, we present a necessary and sufficient condition for Koopman invariance of observables in vRKHSs that are defined via generic non-diagonal operator-valued kernels, and develop sufficient conditions to guarantee the Koopman invariance for vRKHSs defined in terms of a class of diagonal operator-valued kernels. Principles of inverse problems are leveraged to derive error bounds for approximations of the Koopman operator that include both a deterministic sampling error and an approximation error term. The deterministic sampling error arises since imprecisely measured samples are used to approximate the Koopman operator. This work is the first to present overall bounds in the deterministic setting that explicitly account for the sampling error, which, in general, increases with the reduced dimension. Numerical examples illustrate the proposed results. Mathematics Subject Classification (2020) 46E22 · 47A58 · 47B32
- 27 October 2025
- Published by Springer Nature in Research Square
Open AccessDiscrete-Time Dynamical Systems Characterization via Invariance and Approximation of Koopman Operators and Operator-Valued Kernels
This paper presents approaches based on native space theory and Koopman operators to characterize the dynamics of nonlinear, discrete-time dynamical models employing measured data only. Given approximation schemes of the plant dynamics based on Koopman operators contained in vector-valued reproducing kernel Hilbert spaces (vRKHSs), we deduce rates of convergence for these schemes. In particular, we present a necessary and sufficient condition for Koopman invariance of observables in vRKHSs that are defined via generic non-diagonal operator-valued kernels, and develop sufficient conditions to guarantee the Koopman invariance for vRKHSs defined in terms of a class of diagonal operator-valued kernels. Principles of inverse problems are leveraged to derive error bounds for approximations of the Koopman operator that include both a deterministic sampling error and an approximation error term. The deterministic sampling error arises since imprecisely measured samples are used to approximate the Koopman operator. This work is the first to present overall bounds in the deterministic setting that explicitly account for the sampling error, which, in general, increases with the reduced dimension. Numerical examples illustrate the proposed results. Mathematics Subject Classification (2020) 46E22 · 47A58 · 47B32
- Open Access
Tailoring Reproducing Kernels for Optimal Control via Policy Iteration
This paper presents a novel approach to formulating the actor-critic method for optimal control by casting policy iteration in reproducing kernel Hilbert spaces (RKHSs -- also known as native spaces). By tailoring the reproducing kernel and RKHS to the dynamics of the nonlinear optimal control problem, we leverage recent advancements in characterizing error bounds from statistical and machine learning theory. These approximations define a general strategy to select the bases of the actor-critic networks, and we formally guarantee for the first time that this basis selection procedure leads to closed-form error bounds for the individual steps of policy iteration. These bounds often have a geometric and computable form, making them potentially useful for a priori or a posteriori evaluation of candidate collections of scattered bases. Numerical studies subsequently provide qualitative evidence of the practical performance achieved for the full recursion using the algorithms and theory developed for the single-step error bounds.
Open AccessTailoring Reproducing Kernels for Optimal Control via Policy Iteration
This paper presents a novel approach to formulating the actor-critic method for optimal control by casting policy iteration in reproducing kernel Hilbert spaces (RKHSs -- also known as native spaces). By tailoring the reproducing kernel and RKHS to the dynamics of the nonlinear optimal control problem, we leverage recent advancements in characterizing error bounds from statistical and machine learning theory. These approximations define a general strategy to select the bases of the actor-critic networks, and we formally guarantee for the first time that this basis selection procedure leads to closed-form error bounds for the individual steps of policy iteration. These bounds often have a geometric and computable form, making them potentially useful for a priori or a posteriori evaluation of candidate collections of scattered bases. Numerical studies subsequently provide qualitative evidence of the practical performance achieved for the full recursion using the algorithms and theory developed for the single-step error bounds.
- Open Access
MMMnet: A Neural Network Surrogate for Real-Time Transport Prediction Based on the Updated Multi-Mode Model
The Multi-Mode Model (MMM) is a physics-based anomalous transport model integrated into TRANSP for predicting electron and ion thermal transport, electron and impurity particle transport, and toroidal and poloidal momentum transport. While MMM provides valuable predictive capabilities, its computational cost, although manageable for standard simulations, is too high for real-time control applications. MMMnet, a neural network-based surrogate model, is developed to address this challenge by significantly reducing computation time while maintaining high accuracy. Trained on TRANSP simulations of DIII-D discharges, MMMnet incorporates an updated version of MMM (9.0.10) with enhanced physics, including isotopic effects, plasma shaping via effective magnetic shear, unified correlation lengths for ion-scale modes, and a new physics-based model for the electromagnetic electron temperature gradient mode. A key advancement is MMMnet’s ability to predict all six transport coefficients, providing a comprehensive representation of plasma transport dynamics. MMMnet achieves a two-order-of-magnitude speed improvement while maintaining strong correlation with MMM diffusivities, making it well-suited for real-time tokamak control and scenario optimization.
Open AccessMMMnet: A Neural Network Surrogate for Real-Time Transport Prediction Based on the Updated Multi-Mode Model
The Multi-Mode Model (MMM) is a physics-based anomalous transport model integrated into TRANSP for predicting electron and ion thermal transport, electron and impurity particle transport, and toroidal and poloidal momentum transport. While MMM provides valuable predictive capabilities, its computational cost, although manageable for standard simulations, is too high for real-time control applications. MMMnet, a neural network-based surrogate model, is developed to address this challenge by significantly reducing computation time while maintaining high accuracy. Trained on TRANSP simulations of DIII-D discharges, MMMnet incorporates an updated version of MMM (9.0.10) with enhanced physics, including isotopic effects, plasma shaping via effective magnetic shear, unified correlation lengths for ion-scale modes, and a new physics-based model for the electromagnetic electron temperature gradient mode. A key advancement is MMMnet’s ability to predict all six transport coefficients, providing a comprehensive representation of plasma transport dynamics. MMMnet achieves a two-order-of-magnitude speed improvement while maintaining strong correlation with MMM diffusivities, making it well-suited for real-time tokamak control and scenario optimization.
- Open Access
- 7 July 2025
- Published by MDPI AG in Preprints.org
MMMnet: A Neural Network Surrogate for Real-time Transport Prediction Based on the Updated Multi-Mode Model
The Multi-Mode Model (MMM) is a physics-based anomalous transport model integrated into TRANSP for predicting electron and ion thermal transport, electron and impurity particle transport, and toroidal and poloidal momentum transport. While MMM provides valuable predictive capabilities, its computational cost, though manageable for standard simulations, is too high for real-time control applications. MMMnet, a neural network-based surrogate model, is developed to address this challenge by significantly reducing computation time while maintaining high accuracy. Trained on TRANSP simulations of DIII-D discharges, MMMnet incorporates an updated version of MMM (9.0.10) with enhanced physics, including isotopic effects, plasma shaping via effective magnetic shear, unified correlation lengths for ion-scale modes, and a new physics-based model for the electromagnetic electron temperature gradient mode. A key advancement is MMMnet’s ability to predict all six transport coefficients, providing a comprehensive representation of plasma transport dynamics. MMMnet achieves a two-order-of-magnitude speed improvement while maintaining strong correlation with MMM diffusivities, making it well-suited for real-time tokamak control and scenario optimization.
- 7 July 2025
- Published by MDPI AG in Preprints.org
Open AccessMMMnet: A Neural Network Surrogate for Real-time Transport Prediction Based on the Updated Multi-Mode Model
The Multi-Mode Model (MMM) is a physics-based anomalous transport model integrated into TRANSP for predicting electron and ion thermal transport, electron and impurity particle transport, and toroidal and poloidal momentum transport. While MMM provides valuable predictive capabilities, its computational cost, though manageable for standard simulations, is too high for real-time control applications. MMMnet, a neural network-based surrogate model, is developed to address this challenge by significantly reducing computation time while maintaining high accuracy. Trained on TRANSP simulations of DIII-D discharges, MMMnet incorporates an updated version of MMM (9.0.10) with enhanced physics, including isotopic effects, plasma shaping via effective magnetic shear, unified correlation lengths for ion-scale modes, and a new physics-based model for the electromagnetic electron temperature gradient mode. A key advancement is MMMnet’s ability to predict all six transport coefficients, providing a comprehensive representation of plasma transport dynamics. MMMnet achieves a two-order-of-magnitude speed improvement while maintaining strong correlation with MMM diffusivities, making it well-suited for real-time tokamak control and scenario optimization.
- Open Access
- 29 April 2025
- Published by Elsevier in Fusion Engineering and Design
- Vol. 216
Density regulation with disruption avoidance in next-generation tokamaks using a safe reinforcement learning-based controller
- 29 April 2025
- Published by Elsevier in Fusion Engineering and Design
- Vol. 216
Open AccessDensity regulation with disruption avoidance in next-generation tokamaks using a safe reinforcement learning-based controller
- 11 December 2024
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 67  (1) , 015012
Simultaneous control of the electron temperature and safety factor profiles in DIII-D using model-based optimal control techniques
Future tokamak power plants will likely operate using a single, well-defined plasma scenario, either in steady state or for very long pulse lengths. In order to enhance the robustness of the scenario, feedback controllers for a variety of plasma properties will be necessary to counteract any disturbances and ensure safe operation. However, only a limited set of actuators will be available to control many different quantities. Because of this, it is necessary to develop controllers that are able to regulate multiple plasma properties using a limited set of actuators. To this end, a controller has been developed for the simultaneous regulation of both the electron temperature and safety factor profiles in DIII-D. This algorithm uses a linear quadratic integral control synthesis approach based on a linearized model of the dynamics of the two profiles. Two neural network surrogate models, NubeamNet and MMMnet, are included to improve the fidelity of the model. The controller has been tested in simulation using COTSIM, and has demonstrated the ability to simultaneously track changes in both the electron temperature and safety factor targets, including changes in both the magnitude and the shape of the profiles.
- 11 December 2024
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 67  (1) , 015012
Simultaneous control of the electron temperature and safety factor profiles in DIII-D using model-based optimal control techniques
Future tokamak power plants will likely operate using a single, well-defined plasma scenario, either in steady state or for very long pulse lengths. In order to enhance the robustness of the scenario, feedback controllers for a variety of plasma properties will be necessary to counteract any disturbances and ensure safe operation. However, only a limited set of actuators will be available to control many different quantities. Because of this, it is necessary to develop controllers that are able to regulate multiple plasma properties using a limited set of actuators. To this end, a controller has been developed for the simultaneous regulation of both the electron temperature and safety factor profiles in DIII-D. This algorithm uses a linear quadratic integral control synthesis approach based on a linearized model of the dynamics of the two profiles. Two neural network surrogate models, NubeamNet and MMMnet, are included to improve the fidelity of the model. The controller has been tested in simulation using COTSIM, and has demonstrated the ability to simultaneously track changes in both the electron temperature and safety factor targets, including changes in both the magnitude and the shape of the profiles.
- Open Access
- 28 November 2024
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 67  (1) , 015005
Static actuator-sharing algorithm for concurrent control of multiple plasma properties
Simultaneous regulation of multiple properties in next-generation tokamaks like ITER and fusion pilot plant may require the integration of different plasma control algorithms. Such integration requires the conversion of individual controller commands into physical actuator requests while accounting for the coupling between different plasma properties. This work proposes a tokamak and scenario-agnostic actuator-sharing algorithm (ASA) to perform the above-mentioned command-request conversion and, hence, integrate multiple plasma controllers. The proposed algorithm implicitly solves a quadratic programming (QP) problem formulated to account for the saturation limits and the relation between the controller commands and physical actuator requests. Since the constraints arising in the QP program are linear, the proposed ASA is highly computationally efficient and can be implemented in the tokamak plasma control system in real time. Furthermore, the proposed algorithm is designed to handle real-time changes in the control objectives and actuators’ availability. Nonlinear simulations carried out using the Control Oriented Transport SIMulator illustrate the effectiveness of the proposed algorithm in achieving multiple control objectives simultaneously.
- 28 November 2024
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 67  (1) , 015005
Open AccessStatic actuator-sharing algorithm for concurrent control of multiple plasma properties
Simultaneous regulation of multiple properties in next-generation tokamaks like ITER and fusion pilot plant may require the integration of different plasma control algorithms. Such integration requires the conversion of individual controller commands into physical actuator requests while accounting for the coupling between different plasma properties. This work proposes a tokamak and scenario-agnostic actuator-sharing algorithm (ASA) to perform the above-mentioned command-request conversion and, hence, integrate multiple plasma controllers. The proposed algorithm implicitly solves a quadratic programming (QP) problem formulated to account for the saturation limits and the relation between the controller commands and physical actuator requests. Since the constraints arising in the QP program are linear, the proposed ASA is highly computationally efficient and can be implemented in the tokamak plasma control system in real time. Furthermore, the proposed algorithm is designed to handle real-time changes in the control objectives and actuators’ availability. Nonlinear simulations carried out using the Control Oriented Transport SIMulator illustrate the effectiveness of the proposed algorithm in achieving multiple control objectives simultaneously.
- Open Access
Control-Oriented Free-Boundary Equilibrium Solver for Tokamaks
A free-boundary equilibrium solver for an axisymmetric tokamak geometry was developed based on the finite difference method and Picard iteration in a rectangular computational area. The solver can run either in forward mode, where external coil currents are prescribed until the converged magnetic flux function map is achieved, or in inverse mode, where the desired plasma boundary, with or without an X-point, is prescribed to determine the required coil currents. The equilibrium solutions are made consistent with prescribed plasma parameters, such as the total plasma current, poloidal beta, or safety factor at a specified flux surface. To verify the mathematical correctness and accuracy of the solver, the solution obtained using this numerical solver was compared with that from an analytic fixed-boundary equilibrium solver based on the EAST geometry. Additionally, the proposed solver was benchmarked against another numerical solver based on the finite-element and Newton-iteration methods in a triangular-based mesh. Finally, the proposed solver was compared with equilibrium reconstruction results in DIII-D experiments.Open AccessControl-Oriented Free-Boundary Equilibrium Solver for Tokamaks
A free-boundary equilibrium solver for an axisymmetric tokamak geometry was developed based on the finite difference method and Picard iteration in a rectangular computational area. The solver can run either in forward mode, where external coil currents are prescribed until the converged magnetic flux function map is achieved, or in inverse mode, where the desired plasma boundary, with or without an X-point, is prescribed to determine the required coil currents. The equilibrium solutions are made consistent with prescribed plasma parameters, such as the total plasma current, poloidal beta, or safety factor at a specified flux surface. To verify the mathematical correctness and accuracy of the solver, the solution obtained using this numerical solver was compared with that from an analytic fixed-boundary equilibrium solver based on the EAST geometry. Additionally, the proposed solver was benchmarked against another numerical solver based on the finite-element and Newton-iteration methods in a triangular-based mesh. Finally, the proposed solver was compared with equilibrium reconstruction results in DIII-D experiments. - 9 August 2024
- Published by Elsevier in Fusion Engineering and Design
- Vol. 207
Nonlinear control of the minimum safety factor in tokamaks by optimal allocation of spatially moving electron cyclotron current drive
- 9 August 2024
- Published by Elsevier in Fusion Engineering and Design
- Vol. 207
Nonlinear control of the minimum safety factor in tokamaks by optimal allocation of spatially moving electron cyclotron current drive