Keywords
inverter anomaly detection
noisy time series
diffusion reconstruction
signal disentanglement
renewable energy systems
operational fault diagnosis
Abstract
Photovoltaic plants generate multivariate operating time series from inverter voltage, current, power output, temperature, string-level generation, irradiance, grid frequency, and alarm states. These signals are noisy because of cloud movement, shading, sensor drift, weather fluctuations, and grid-side disturbances. Conventional anomaly detection methods may confuse normal irradiance-induced variation with inverter degradation or electrical faults. This study develops a noise-resilient disentanglement method for photovoltaic inverter anomaly detection. The method first applies conditional diffusion reconstruction to denoise operating signals under changing irradiance and temperature conditions. A factor-disentanglement module then separates weather-driven variation, grid fluctuation, and inverter-related abnormal components. An adaptive residual boundary is used to detect anomalies from denoised inverter-specific representations. Experiments are conducted on a photovoltaic plant dataset containing 96 solar farms, 12,800 inverters, 284,000 string channels, and 42 operating variables collected at one-minute intervals over 20 months. The dataset contains 1.34 billion timestamped records and 3,260 confirmed abnormal events, including inverter overheating, string current mismatch, DC-side insulation decline, abnormal clipping, grid-frequency disturbance, and MPPT tracking failure. The proposed method shortens median detection delay from 47.5 minutes to 13.2 minutes compared with a temporal reconstruction baseline. False alerts decrease to 2.4 cases per solar farm per month. The denoised current and voltage trajectories show a mean reconstruction deviation of 0.064 normalized units, while weather-related components absorb 71 million cloud-transient windows before anomaly scoring. The online scoring module processes 64,000 inverter records per second with a median latency of 33 ms per farm window. These findings demonstrate that denoising and disentangled representation learning can improve anomaly detection in noisy photovoltaic operation time series.
References
Qi, C., & Qiao, X. (2026). Building Reliable AI Systems: A Framework That Bridges Architecture and Operations. Available at SSRN 6795298.
Hassan, O., Yusupov, Z., & Abubakr, M. M. (2026). Hybrid AI and Optimization Algorithms for Performance Enhancement of Grid-Connected Solar PV Systems. Int. J. Electr. Eng. Sustain., 4(1), 63-84.
Hong, Z., Xu, T., & Chen, H. (2026). Automated Detection and Reclamation of Orphaned Compute and Storage Resources in Cloud Infrastructure. Available at SSRN 7116258.
Sharma, P., & Yadav, A. (2027). A hybrid deep learning and adaptive residual framework for accurate solar power forecasting and PV system reliability enhancement. Electric Power Systems Research, 264, 113854.
Chen, X., Xiao, H., Zeng, Z., Zhang, S., & Xiao, R. (2025). Fine-Grained Multivariate Time Series Anomaly Detection via Causal Inference. Knowledge-Based Systems, 114765.
Fingerhut, F., & Tsiporkova, E. (2026). Multi-view Data Exploration Recipes for Making Sense of Multi-source Industrial Data. Ieee Access.
Li, J., Ma, R., & Gu, Y. (2026). From Audit Requirements to Computable Software Architecture: A Rule-Engine and Evidence-Indexing Method for Trusted Digital Infrastructure.
Qi, C., & Qiao, X. (2026). Efficient Data Sampling and Feature Selection Algorithms for Scalable Machine Learning Pipelines.
Zhang, Y., Gu, W., & Wang, J. (2025, December). Research on First Article Inspection (FAI)-Driven Quality Assurance Methods for Wind Turbine Installation and Operation & Maintenance and Their Effect on Reliability Improvement. In Proceedings of the 2025 6th International Conference on Computer Science and Management Technology (pp. 1700-1705).
Angulo, J. R., Conde, L. A., Pleshcheva, V., García, M. A., Gosgot, W., Barrena, M., ... & Töfflinger, J. A. (2025). Effective DC power rating of PV arrays under challenging operating conditions in desert and tropical regions. Renewable Energy, 124981.
Su, D., & Dong, Y. (2026). Classroom-Based Assessment with Bayesian Learning Analytics for Instructional Decision-Making in ASD Inclusive Education.
Xu, T., Zhu, W., & Zhang, J. (2026). Stability and Consistency of Explainable Deep Learning Methods in Credit Risk Assessment. Available at SSRN 6893938.
Kage, P., Rothenberger, J., Andreadis, P., & Diochnos, D. (2026). A review of pseudo-labeling for computer vision. Journal of Artificial Intelligence Research, 85.
Du, Y., Dong, Y., Liu, Q., & Chen, Z. (2026). Modal Reliability Assessment and Drift Early Warning in Visible-Infrared Target Tracking. Available at SSRN 7132381.
Zhang, Z., Tong, Y., & Gao, Y. (2026). Retrieval-Augmented Generation with Low-Latency Deployment for Vertical Domains Question Answering: A Case Study on Economic Resource Platforms.
Bustamante, P. (2026). Bayesian Deep Learning for Photovoltaic Power Forecasting: Probabilistic Framework for Uncertainty Quantification and Grid Reliability Optimization (Doctoral dissertation, The University of Texas at El Paso).
Zhang, Y., & Wang, J. (2026, February). Design and Implementation of a Computer-Aided Full Lifecycle Quality Management System for Wind Farms in Upgrades, Renovations, and Subcontractor Supervision. In Proceedings of the 2026 International Conference on Big Data and Informatization Education (pp. 1309-1314).
Su, D., & Zhang, H. Developing a Data-Driven Computational Framework for Scalable Special Education Practices for Autism and Intellectual Disabilities in the US Public School System.
Xiong, W., & Zeng, Y. (2026). Multi-Layer Coupled Diagnosis of Energy Efficiency Degradation in Complex Building MEP Systems and Identification of Optimization Boundaries. Available at SSRN 7121782.
Kull, K., Asad, B., Khan, M. A., Naseer, M. U., Kallaste, A., & Vaimann, T. (2025). Faults, failures, reliability, and predictive maintenance of grid-connected solar systems: A comprehensive review. Applied Sciences, 15(21), 11461.
Solís-Martín, D., Galán-Páez, J., & Borrego-Díaz, J. (2026). D3a-ts: Denoising-driven data augmentation in time series. International Journal of Data Science and Analytics, 22(1), 48.
Hong, Z., Chen, H., & Xu, T. (2026). Long-Horizon Load Stability Testing and Degradation Forecasting for High-Concurrency Cloud Storage. Available at SSRN 7115859.
Li, J., Zhang, Y., & Gu, Y. (2026). Chain-Hash Audit Framework with Trusted Anchoring for Tamper-Evident Evidence Lifecycle Management.
Ganguli, A., Ramachandra, N., Bessac, J., & Constantinescu, E. (2025). Enhancing interpretability in generative modeling: statistically disentangled latent spaces guided by generative factors in scientific datasets. Machine Learning, 114(9), 197.
Du, Y., Chen, Z., Liu, Q., & Dong, Y. (2026). Co-identification of Vibration, Structural Looseness, and Surface Damage in Industrial Equipment Operation Videos.
Xiong, W., Zeng, Y., & Guo, Y. (2026). A Study on the Characterization of Building MEP System States and the Learning of Behavioral Patterns Driven by Large-Scale Operational Data. Available at SSRN 7121883.
Guanghua, L., Shah, S. H. H., Zaidi, A. A., Nawaz, F., & Khan, S. Z. (2026). A Review of State‐of‐the‐Art Developments in Large‐Scale Photovoltaic Solar Renewable Energy Systems. Journal of Engineering, 2026(1), 4799038.
Fu, Y., Gui, H., Li, W., & Wang, Z. (2020, August). Virtual Material Modeling and Vibration Reduction Design of Electron Beam Imaging System. In 2020 IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA) (pp. 1063-1070). IEEE.
Yin, J., Huang, Y., & Rao, H. (2026). A Study on the Dynamic Evolution of Learning Behavior and Outcome Prediction in Digital Educational Environments. Available at SSRN 6607139.
Kijanowski, K., Barszcz, T., & Dao, P. B. (2025). A cluster-based filtering approach to SCADA data preprocessing for wind turbine condition monitoring and fault detection. Energies, 18(22), 5954.
Kull, K., Asad, B., Khan, M. A., Naseer, M. U., Kallaste, A., & Vaimann, T. (2025). Faults, failures, reliability, and predictive maintenance of grid-connected solar systems: A comprehensive review. Applied Sciences, 15(21), 11461.
Zhang, Z. (2026). A Study on Multimodal Product Understanding and Assembly Guidance Optimization in e-commerce for Complex Consumer Goods. Available at SSRN 6734559.