Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

IEEE Holm Conference on Electrical Contacts (HOLM)

Variational Quantum Neural Network for Modeling and Solving Heat and Mass Transfer Problems Occurring in Electric Contact Phenomena

Merey M. Sarsengeldin, Zeeshan Ahmad, Ulaş Bağcı

Published Oct 6, 2024• 1 citation
Publisher

Abstract

In this study we present a hybrid quantum-classical neural network (variational quantum Algorithm) to model and solve heat and mass transfer problems that occur in electrical contacts. The underlying Equations responsible for modeling diverse electric contact phenomena like arcing, bridging, welding etc. are Stefan Type Problems. These problems are nonlinear and hard to solve analytically. We build a Hybrid Quantum-Classical Neural Network for modeling and solving such kind Heat and Mass transfer problems occurring in Electric Contacts. We benefit from computational power of Quantum Computing to solve the Stefan Type problems and utilize neural network as a universal function approximator for the Heat function and Moving Phase boundary. On the basis of the Variational Quantum Neural Network, we have developed methodological framework and software artifact which might be of interest and beneficial for researchers and engineers working in the field of modeling electric contact phenomena. The proposed framework and artifact are beyond the field of modeling and solving Heat and Mass transfer problems occurring in electric contacts and can be applied in diverse phase transient phenomena such as tumor growth and spreading (biomedicine), material deformation and failure under stress (material science) etc. Methodology and source code for the framework is publicly available on our GitHub account https://github.com/Schrodinger-cat-kz/QNN-Framework-for-Electric-Contact-Phenomena.git.

DOI
10.1109/holm56222.2024.10768556

Related papers

Special Functions and HHL Quantum Algorithm for Solving Moving Boundary Value Problems Occurring in Electric Contact Phenomena

Merey M. Sarsengeldin, M. Zuhair Nashed

This series of studies is devoted to developing and employing mathematical methods along with quantum algorithms for solving moving boundary value problems which occur in heat and mass transfer problems. In this particular study we develop mathematical framework where we utilize special functions and Harrow-Hassidim-Lloyd (HHL) quantum algorithm for finding exact and approximate solutions of Generalized Heat Equation with moving boundaries and as examples we consider plane and spherical cases. In spherical case the Generalized Heat Equation is reduced to linear moving boundary value problem with discontinuous coefficients and solved exactly. In plane case we use collocation method for approximate solution of Inverse Two-Phase Stefan problem. Experimental verification of suggested mathematical framework has been tested for modeling arcing phenomena in composite electrical contacts with AgCdO (90%) and Ni (10%). HHL algorithm was applied and run on IBM Q with Qiskit and the code is openly available on https://github.com/users/Schrodinger-cat-kz/projects/2.

arXiv:2010.05691•IEEE Holm Conference on Electrical Contacts (HOLM)•Oct 24, 2021•0 citations

A Hybrid Classical-Quantum framework for solving Free Boundary Value Problems and Applications in Modeling Electric Contact Phenomena

Merey M. Sarsengeldin

In this paper we elaborate a hybrid classical-quantum framework which allows one to model and solve heat and mass transfer problems occurring in electric contacts. We utilize special functions and Harrow-Hassidim-Lloyd (HHL) quantum algorithm for finding temperature and arc flux functions exactly and approximately for the Stefan type problems. The Stefan type problems we are considering are based on the Generalized Heat Equation with free boundaries. As examples we consider exact and approximate solutions of inverse one-phase and two-phase Stefan problems. An Inverse Generalized One-Phase Stefan Problem is considered as a model problem. Computational experiments were conducted and demonstrated on IBM Quantum Machine.

arXiv:2205.02230•May 4, 2022•1 citation

Solving nonlinear PDEs with Quantum Neural Networks: A variational approach to the Bratu Equation

Nikolaos Cheimarios

We present a variational quantum algorithm (VQA) to solve the nonlinear one-dimensional Bratu equation. By formulating the boundary value problem within a variational framework and encoding the solution in a parameterized quantum neural network (QNN), the problem reduces to an optimization task over quantum circuit parameters. The trial solution incorporates a predictor from the previous continuation step and boundary-enforcing terms, allowing the circuit to focus on minimizing the residual of the differential operator. Using a noiseless quantum simulator, we demonstrate that the method accurately captures both solution branches of the Bratu equation and shows excellent agreement with classical pseudo arc-length continuation results.

quant-pharXiv:2601.04372•Jan 9, 2026•1 citation

Variational Quantum Neural Networks (VQNNS) in Image Classification

Meghashrita Das, Tirupati Bolisetti

Quantum machine learning has established as an interdisciplinary field to overcome limitations of classical machine learning and neural networks. This is a field of research which can prove that quantum computers are able to solve problems with complex correlations between inputs that can be hard for classical computers. This suggests that learning models made on quantum computers may be more powerful for applications, potentially faster computation and better generalization on less data. The objective of this paper is to investigate how training of quantum neural network (QNNs) can be done using quantum optimization algorithms for improving the performance and time complexity of QNNs. A classical neural network can be partially quantized to create a hybrid quantum-classical neural network which is used mainly in classification and image recognition. In this paper, a QNN structure is made where a variational parameterized circuit is incorporated as an input layer named as Variational Quantum Neural Network (VQNNs). We encode the cost function of QNNs onto relative phases of a superposition state in the Hilbert space of the network parameters. The parameters are tuned with an iterative quantum approximate optimisation (QAOA) mixer and problem hamiltonians. VQNNs is experimented with MNIST digit recognition (less complex) and crack image classification datasets (more complex) which converges the computation in lesser time than QNN with decent training accuracy.

arXiv:2303.05860•Mar 10, 2023•19 citations

Toward Variational Quantum Algorithms for Generalized Linear and Nonlinear Transport Phenomena

Sergio Bengoechea, Paul Over, Dieter Jaksch +1 more

This article proposes a variational quantum algorithm to solve linear and nonlinear thermofluid dynamic transport equations. The hybrid classical-quantum framework is applied to problems governed by the heat, wave, and Burgers’ equations in combination with different engineering boundary conditions. Topics covered include the encoding of band matrices, as in the consideration of nonconstant material properties and upwind-biased first- and higher-order approximations, widely used in engineering computational fluid dynamics, by the use of a mask function. Verification examples demonstrate high predictive agreement with classical methods. Furthermore, the scalability analysis shows a polylog scaling of the number of quantum gates with the number of qubits. Remaining challenges refer to the implicit construction of upwind schemes and the identification of an appropriate parameterization strategy of the quantum ansatz.

AIAA Journal•Nov 5, 2025•12 citations

A Quantum Computing-Based Neural Network Approach for Modeling Electrical Discharge Machining

A. Karunyan, Raja Das

Electrical discharge machining (EDM) is a stochastic procedure used for machining difficultto- cut materials such as AISI D2 steel; yet, precisely forecasting surface roughness (Ra) is problematic due to the non-linear interactions among the process parameters. Although data-driven models like the Backpropagation Neural Networks (BPNNs) are extensively utilized, enhancing their generalization ability on a constrained experimental dataset continues to be a prominent research focus. This study assesses a hybrid Quantum Neural Network (QNN) framework for predictive modelling. The architecture utilizes Angle Encoding through parameterized Rx rotation gates to transform classical inputs such as (Peak Current, Pulse-on Time, Duty Factor, Voltage) into quantum states, subsequently employing a variational circuit with linear entanglement to capture complex feature correlations. The suggested model was executed and verified in the MATLAB environment to illustrate the algorithmic benefits of the quantum processes. Experimental validation was performed utilizing a dataset including 90 observations obtained from the machining trials. The QNN’s performance was evaluated in comparison to standard Artificial Neural Networks (ANN), Radial Basis Function Networks (RBFN), and Support Vector Regression (SVR). The QNNmodel attained a Correlation Coefficient (R) of 0.9643 and a Mean Squared Error (MSE) of 0.1642. The results demonstrate that the suggested quantum architecture provides enhanced predictive accuracy for this application, outperforming the conventional classical baselines in modelling the intricate surface roughness profile.

IEEE Access•Jan 1, 2026•0 citations
In this study we present a hybrid quantum-classical neural network (variational quantum Algorithm) to model and solve heat and mass transfer problems that occur in electrical contacts. The underlying Equations responsible for modeling diverse electric contact phenomena like arcing, bridging, welding etc. are Stefan Type Problems. These problems are nonlinear and hard to solve analytically. We build a Hybrid Quantum-Classical Neural Network for modeling and solving such kind Heat and Mass transfer problems occurring in Electric Contacts. We benefit from computational power of Quantum Computing to solve the Stefan Type problems and utilize neural network as a universal function approximator for the Heat function and Moving Phase boundary. On the basis of the Variational Quantum Neural Network, we have developed methodological framework and software artifact which might be of interest and beneficial for researchers and engineers working in the field of modeling electric contact phenomena. The proposed framework and artifact are beyond the field of modeling and solving Heat and Mass transfer problems occurring in electric contacts and can be applied in diverse phase transient phenomena such as tumor growth and spreading (biomedicine), material deformation and failure under stress (material science) etc. Methodology and source code for the framework is publicly available on our GitHub account https://github.com/Schrodinger-cat-kz/QNN-Framework-for-Electric-Contact-Phenomena.git.
This series of studies is devoted to developing and employing mathematical methods along with quantum algorithms for solving moving boundary value problems which occur in heat and mass transfer problems. In this particular study we develop mathematical framework where we utilize special functions and Harrow-Hassidim-Lloyd (HHL) quantum algorithm for finding exact and approximate solutions of Generalized Heat Equation with moving boundaries and as examples we consider plane and spherical cases. In spherical case the Generalized Heat Equation is reduced to linear moving boundary value problem with discontinuous coefficients and solved exactly. In plane case we use collocation method for approximate solution of Inverse Two-Phase Stefan problem. Experimental verification of suggested mathematical framework has been tested for modeling arcing phenomena in composite electrical contacts with AgCdO (90%) and Ni (10%). HHL algorithm was applied and run on IBM Q with Qiskit and the code is openly available on https://github.com/users/Schrodinger-cat-kz/projects/2.
In this paper we elaborate a hybrid classical-quantum framework which allows one to model and solve heat and mass transfer problems occurring in electric contacts. We utilize special functions and Harrow-Hassidim-Lloyd (HHL) quantum algorithm for finding temperature and arc flux functions exactly and approximately for the Stefan type problems. The Stefan type problems we are considering are based on the Generalized Heat Equation with free boundaries. As examples we consider exact and approximate solutions of inverse one-phase and two-phase Stefan problems. An Inverse Generalized One-Phase Stefan Problem is considered as a model problem. Computational experiments were conducted and demonstrated on IBM Quantum Machine.
We present a variational quantum algorithm (VQA) to solve the nonlinear one-dimensional Bratu equation. By formulating the boundary value problem within a variational framework and encoding the solution in a parameterized quantum neural network (QNN), the problem reduces to an optimization task over quantum circuit parameters. The trial solution incorporates a predictor from the previous continuation step and boundary-enforcing terms, allowing the circuit to focus on minimizing the residual of the differential operator. Using a noiseless quantum simulator, we demonstrate that the method accurately captures both solution branches of the Bratu equation and shows excellent agreement with classical pseudo arc-length continuation results.
Quantum machine learning has established as an interdisciplinary field to overcome limitations of classical machine learning and neural networks. This is a field of research which can prove that quantum computers are able to solve problems with complex correlations between inputs that can be hard for classical computers. This suggests that learning models made on quantum computers may be more powerful for applications, potentially faster computation and better generalization on less data. The objective of this paper is to investigate how training of quantum neural network (QNNs) can be done using quantum optimization algorithms for improving the performance and time complexity of QNNs. A classical neural network can be partially quantized to create a hybrid quantum-classical neural network which is used mainly in classification and image recognition. In this paper, a QNN structure is made where a variational parameterized circuit is incorporated as an input layer named as Variational Quantum Neural Network (VQNNs). We encode the cost function of QNNs onto relative phases of a superposition state in the Hilbert space of the network parameters. The parameters are tuned with an iterative quantum approximate optimisation (QAOA) mixer and problem hamiltonians. VQNNs is experimented with MNIST digit recognition (less complex) and crack image classification datasets (more complex) which converges the computation in lesser time than QNN with decent training accuracy.
This article proposes a variational quantum algorithm to solve linear and nonlinear thermofluid dynamic transport equations. The hybrid classical-quantum framework is applied to problems governed by the heat, wave, and Burgers’ equations in combination with different engineering boundary conditions. Topics covered include the encoding of band matrices, as in the consideration of nonconstant material properties and upwind-biased first- and higher-order approximations, widely used in engineering computational fluid dynamics, by the use of a mask function. Verification examples demonstrate high predictive agreement with classical methods. Furthermore, the scalability analysis shows a polylog scaling of the number of quantum gates with the number of qubits. Remaining challenges refer to the implicit construction of upwind schemes and the identification of an appropriate parameterization strategy of the quantum ansatz.
Electrical discharge machining (EDM) is a stochastic procedure used for machining difficultto- cut materials such as AISI D2 steel; yet, precisely forecasting surface roughness (Ra) is problematic due to the non-linear interactions among the process parameters. Although data-driven models like the Backpropagation Neural Networks (BPNNs) are extensively utilized, enhancing their generalization ability on a constrained experimental dataset continues to be a prominent research focus. This study assesses a hybrid Quantum Neural Network (QNN) framework for predictive modelling. The architecture utilizes Angle Encoding through parameterized Rx rotation gates to transform classical inputs such as (Peak Current, Pulse-on Time, Duty Factor, Voltage) into quantum states, subsequently employing a variational circuit with linear entanglement to capture complex feature correlations. The suggested model was executed and verified in the MATLAB environment to illustrate the algorithmic benefits of the quantum processes. Experimental validation was performed utilizing a dataset including 90 observations obtained from the machining trials. The QNN’s performance was evaluated in comparison to standard Artificial Neural Networks (ANN), Radial Basis Function Networks (RBFN), and Support Vector Regression (SVR). The QNNmodel attained a Correlation Coefficient (R) of 0.9643 and a Mean Squared Error (MSE) of 0.1642. The results demonstrate that the suggested quantum architecture provides enhanced predictive accuracy for this application, outperforming the conventional classical baselines in modelling the intricate surface roughness profile.