PiGraphNet: A Novel Physics-Informed Graph Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries in Electric Vehicles Manoranjan Gandhudi, Gangadharan G. R 2025 IEEE 4th International Conference on Smart Technologies for Power Energy and Control Stpec 2025 Conference Report, 2025 Accurate prediction of the Remaining Useful Life of lithium-ion batteries is essential for enhancing the performance, safety, and reliability of electric vehicles (EVs). Existing model-based approaches often require extensive parameter identification and lose accuracy under unmodeled operating conditions, whereas purely data-driven methods may overlook spatial-temporal dependencies and violate physical constraints. To address these limitations, we propose PiGraphNet, a novel Physics-Informed Graph Neural Network (GNN) that unifies graph-based spatial-temporal modeling with embedded battery degradation physics. PiGraphNet dynamically constructs battery interaction graphs from voltage, current, and temperature data to capture evolving inter-cell dependencies, while integrating electrochemical degradation laws, such as Peukert's law, into the message-passing mechanism for physically consistent predictions. Two domain-specific graph strategies—the Voltage-Current-Temperature (VCT) Graph and the Electrode Degradation Graph (EDG) enable targeted modeling of real-time electro-thermal coupling and long-term aging patterns. The model is trained with a composite loss function combining mean squared error with physics-informed constraints, ensuring both accuracy and interpretability. Experimental results on the NASA lithium-ion battery degradation dataset demonstrate that PiGraphNet outperforms LSTM, GRU, TCN, and physics-informed baselines, achieving an MSE of 251.60 and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$R^{2}$</tex> of 93.29%.
Q-ForecasterNet: A Hybrid Quantum Neural Network for Sales Forecasting Manoranjan Gandhudi, Alphonse P J A, Gangadharan G R 2025 17th IEEE International Conference on Computational Intelligence and Communication Networks Cicn 2025, 2025 Sales forecasting is crucial for guiding strategic decisions in retail businesses, enabling effective inventory management, staffing, and promotional planning. Accurate forecasting helps minimize losses due to overstocking or stockouts and supports data-driven operational efficiency. Recently, machine learning (ML) techniques have been adopted for sales prediction, which are capable of identifying complex hidden patterns of data from historical organizational data. However, traditional ML models often struggle to analyze longterm dependencies and handle dynamic, nonlinear relationships within time-series of organizational data. To overcome these difficulties, we introduce a Q-ForecasterNet model which integrates a temporal convolutional network encoder, a long short-term memory layer for sequential memory, and a variational quantum circuit to embed decision-relevant representations. Experiments on the Rossmann store sales dataset present that the proposed model significantly outperforms classical and deep learning baselines, achieving the optimal prediction errors with an MSE of 126.9, RMSE of 11.26, MAE of 1.12, and MAPE of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 4 8}$</tex>.
Causal Inference and Explainable Artificial Intelligence based Quantum Deep Learning for Remaining Useful Lifetime Prediction M Gandhudi, PJA Alphonse, S Srinivas, GR Gangadharan Knowledge-Based Systems, 115669 , 2026 2026 Citations: 1
Dynamic quantum annealing optimized quantum neural networks for remaining useful lifetime prediction M Gandhudi, GR Gangadharan Engineering Applications of Artificial Intelligence 167, 113856 , 2026 2026
Q-ForecasterNet: A Hybrid Quantum Neural Network for Sales Forecasting M Gandhudi, PJA Alphonse, GR Gangadharan 2025 IEEE 17th International Conference on Computational Intelligence and … , 2025 2025
PiGraphNet: A Novel Physics-Informed Graph Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries in Electric Vehicles M Gandhudi 2025 IEEE 4th International Conference on Smart Technologies for Power … , 2025 2025
Quantum machine learning for stock price movement prediction M Gandhudi, PJA Alphonse, GR Gangadharan Data Analytics for Decision Making towards Business Excellence: Through the … , 2025 2025 Citations: 1
Dynamic customer behavior prediction in subscription services using causal reinforcement learning M Gandhudi, PJA Alphonse, V Velayudham, L Nagineni, ... Engineering Applications of Artificial Intelligence 155, 111030 , 2025 2025 Citations: 7
VWFTS-PSO: a novel method for time series forecasting using variational weighted fuzzy time series and particle swarm optimization G Didugu, M Gandhudi, PJA Alphonse, GR Gangadharan International Journal of General Systems 54 (4), 540-559 , 2025 2025 Citations: 9
Explainable causal variational autoencoders based equivariant graph neural networks for analyzing the consumer purchase behavior in E-commerce M Gandhudi, PJA Alphonse, V Velayudham, L Nagineni, ... Engineering Applications of Artificial Intelligence 136, 108988 , 2024 2024 Citations: 34
Explainable hybrid quantum neural networks for analyzing the influence of tweets on stock price prediction M Gandhudi, PJA Alphonse, U Fiore, GR Gangadharan Computers and Electrical Engineering 118, 109302 , 2024 2024 Citations: 37
AB-bil: A deep learning model to analyze depression detection in imbalanced data RK Bondugula, M Gandhudi, KB Sivangi, J Ahamed, MA Chaurasia, ... Smart Healthcare and Machine Learning, 9-16 , 2024 2024 Citations: 1
Causal aware parameterized quantum stochastic gradient descent for analyzing marketing advertisements and sales forecasting M Gandhudi, GR Gangadharan, PJA Alphonse, V Velayudham, ... Information Processing & Management 60 (5), 103473 , 2023 2023 Citations: 20
Security against ssdf attacks using novel attack mitigation mechanism for cognitive radio networks G Manoranjan, M Chandan, G Karthik, R Satpathy, S Sugumaran 2021 Fourth International Conference on Electrical, Computer and … , 2021 2021 Citations: 6
Service Composition in Mobile Ad Hoc Networks (MANET’s) with the Help of Optimal QoS Constraints G Manoranjan, MV Rathnamma, V Venkata Ramana, GR Anil International Conference On Computational And Bio Engineering, 375-390 , 2019 2019 Citations: 3
MOST CITED SCHOLAR PUBLICATIONS
Explainable hybrid quantum neural networks for analyzing the influence of tweets on stock price prediction M Gandhudi, PJA Alphonse, U Fiore, GR Gangadharan Computers and Electrical Engineering 118, 109302 , 2024 2024 Citations: 37
Explainable causal variational autoencoders based equivariant graph neural networks for analyzing the consumer purchase behavior in E-commerce M Gandhudi, PJA Alphonse, V Velayudham, L Nagineni, ... Engineering Applications of Artificial Intelligence 136, 108988 , 2024 2024 Citations: 34
Causal aware parameterized quantum stochastic gradient descent for analyzing marketing advertisements and sales forecasting M Gandhudi, GR Gangadharan, PJA Alphonse, V Velayudham, ... Information Processing & Management 60 (5), 103473 , 2023 2023 Citations: 20
VWFTS-PSO: a novel method for time series forecasting using variational weighted fuzzy time series and particle swarm optimization G Didugu, M Gandhudi, PJA Alphonse, GR Gangadharan International Journal of General Systems 54 (4), 540-559 , 2025 2025 Citations: 9
Dynamic customer behavior prediction in subscription services using causal reinforcement learning M Gandhudi, PJA Alphonse, V Velayudham, L Nagineni, ... Engineering Applications of Artificial Intelligence 155, 111030 , 2025 2025 Citations: 7
Security against ssdf attacks using novel attack mitigation mechanism for cognitive radio networks G Manoranjan, M Chandan, G Karthik, R Satpathy, S Sugumaran 2021 Fourth International Conference on Electrical, Computer and … , 2021 2021 Citations: 6
Service Composition in Mobile Ad Hoc Networks (MANET’s) with the Help of Optimal QoS Constraints G Manoranjan, MV Rathnamma, V Venkata Ramana, GR Anil International Conference On Computational And Bio Engineering, 375-390 , 2019 2019 Citations: 3
Causal Inference and Explainable Artificial Intelligence based Quantum Deep Learning for Remaining Useful Lifetime Prediction M Gandhudi, PJA Alphonse, S Srinivas, GR Gangadharan Knowledge-Based Systems, 115669 , 2026 2026 Citations: 1
Quantum machine learning for stock price movement prediction M Gandhudi, PJA Alphonse, GR Gangadharan Data Analytics for Decision Making towards Business Excellence: Through the … , 2025 2025 Citations: 1
AB-bil: A deep learning model to analyze depression detection in imbalanced data RK Bondugula, M Gandhudi, KB Sivangi, J Ahamed, MA Chaurasia, ... Smart Healthcare and Machine Learning, 9-16 , 2024 2024 Citations: 1
Dynamic quantum annealing optimized quantum neural networks for remaining useful lifetime prediction M Gandhudi, GR Gangadharan Engineering Applications of Artificial Intelligence 167, 113856 , 2026 2026
Q-ForecasterNet: A Hybrid Quantum Neural Network for Sales Forecasting M Gandhudi, PJA Alphonse, GR Gangadharan 2025 IEEE 17th International Conference on Computational Intelligence and … , 2025 2025
PiGraphNet: A Novel Physics-Informed Graph Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries in Electric Vehicles M Gandhudi 2025 IEEE 4th International Conference on Smart Technologies for Power … , 2025 2025
