A Hybrid Approach to Probabilistic Solar Forecasting
The solar industry needs more accurate solar forecasting which is why the U.S. Department of Energy Solar Energy Technologies Office sponsored the American-Made Solar Forecasting Prize. This competition was designed to incentivize the advancement of probabilistic methods for day ahead models that incorporated weather-related uncertainties such as cloud coverage. One of the Prize winners announced in March 2022 was Dr. Wenyuan Tang, an Assistant Professor in Electrical and Computer Engineering and FREEDM researcher. He proposed a hybrid approach to probabilistic forecasting that has low data cost and low computational costs. It includes learning from many base models, as well as physical and statistical models, to provide a comprehensive tool for both forecasting and grid optimization. This model is available for licensing. More technical details are provided in the references below.
- Junkai Liang and Wenyuan Tang, “Ultra-short-term spatiotemporal forecasting of renewable resources: An attention temporal convolutional network based approach,” IEEE Transactions on Smart Grid, vol. 13, no. 5, pp. 3798–3812, 2022.
- Ashwin Shirsat and Wenyuan Tang, “Data-driven stochastic model predictive control for DC-coupled residential PV-storage systems,” IEEE Transactions on Energy Conversion, vol. 36, no. 2, pp. 1435–1448, 2021.
- Junkai Liang and Wenyuan Tang, “Scenario reduction for stochastic day-ahead scheduling: A mixed autoencoder based time-series clustering approach,” IEEE Transactions on Smart Grid, vol. 12, no. 3, pp. 2652–2662, 2021.
- Junkai Liang and Wenyuan Tang, “Interval based transmission contingency-constrained unit commitment for integrated energy systems with high renewable penetration,” International Journal of Electrical Power & Energy Systems, vol. 119, 2020, 105853.
- Junkai Liang, Ashwin Shirsat, and Wenyuan Tang, “Sustainable community based PV-storage planning using the Nash bargaining solution,” International Journal of Electrical Power & Energy Systems, vol. 118, 2020, 105759.
- Junkai Liang and Wenyuan Tang, “Sequence generative adversarial networks for wind power scenario generation,” IEEE Journal on Selected Areas in Communications, vol. 38, no. 1, pp. 110–118, 2020.
Utility Scale PV Power Optimization and Prediction Software Package
NCSU researchers led by Dr. Ning Lu created two software applications that dramatically increase utility scale PV power stability and increase accuracy and resolution of predicted power output. PV forecasting is improved through Temporal Convolutional Networks applied to historical power production from nearby PV farms. Power regulation applies a novel Perturb and Observe technique to achieve faster set point convergence than existing methods. Together, these applications lead to higher revenue for system operators.
Novel Inverter Topology is Finalist for Solar Prize
A FREEDM spinoff, NC Solar Inverters, is a finalist for Round 6 of the American Made Solar Prize. Ken Dulaney and Wensong Yu formed the company after licensing technology for a new type of solar inverter that uses complex controls to combine high frequency switching devices with low frequency switching devices for reliable performance at lower cost and lower losses than today’s state of the art for solar inverters. The team is a Finalist for the Go! stage of the competition and will present their 50 kW prototype and business plan at RE+ 2023 in Las Vegas in September.