Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems

Significance 

With the increasing demand and usage of photovoltaic (PV) and wind power, the inherent uncertainties of these power sources have continued to compromise the operation stability of various power systems. Some methods used to model the uncertainties to address the impact of photovoltaic and wind power include fuzzy optimization, distributionally robust optimization and stochastic optimization. Nevertheless, despite the significant efforts, accurate description of the uncertainties has remained a great challenge. This can be partly attributed to the fact that the uncertainties follow unknown distributions and are usually corrupted by anomalous measurements. Additionally, the extraction of distributional information and accurate intervals from the uncertainties is challenging.

In addition to renewable energy, the development of nuclear power and its subsequent connection to the grid has rapidly grown recently, further complicating the operation of power systems. It also leads to numerous contradictions amongst the various power sources and can lead to curtailments of wind power and water spillage (releasing water without generation of electricity). To this end, it is also necessary to establish effective models for nuclear power plants (NPPs). However, despite the improved research on multi-source optimal scheduling considering pick load regulation, only a few of them involve nuclear power. Notably, water spoilage is a key problem associated with the generation of hydroelectric power. Its optimization, which currently involves the optimization of the entire basin, has proved insufficient. Therefore, the development of new water spillage strategies is also indispensable.

In light of the above, Dr. Wenting Hou from Zhoukou Normal University and Professor Hua Wei from Guangxi University proposed a data-driven robust day-ahead unit commitment optimization model for hydro-thermal-wind-PV-nuclear power systems. The method is based on the robust kernel density estimation (RKDE) and was used to model PV and wind power uncertainties. The power generation plan was created by reducing the water spillages in hydro stations, operation costs of the thermal units, and the peak shaving cost of the nuclear plants. The work is currently published in the journal, Electrical Power and Energy Systems.

In their approach, the authors focused on the optimal operation of the different power sources. The RKDE extracted the distribution data from the big data of the PV and wind power. Together with the quantile functions, the distribution information was used to develop the data-driven framework. A total of eight water spillage strategies were considered. The water spillages were distributed based on four main principles: overall optimization of the entire basin, proportionally to the installed capacity, proportion to the annual completion rate of the plan, and proportion to user-defined ratio. Overall, the proposed new framework aimed to minimize operation costs, water spillages and enhance operational flexibility.

The results of the simulations conducted on a modified New England 39-bus system demonstrated the flexibility, practicability and superiority of the model. The eight strategies could achieve the allocation of water spillage losses in hydro plants, thereby realizing the desirable balance between the individual hydropower plants and the overall basin. In addition, the peak shaving depth and low-power operating time of the NPPs were determined at the rescheduling stage and pre-scheduling stage, respectively. The model could make use of the complementary characteristics of different power sources and effectively coordinate their operations. Furthermore, since the water spillages strategies have various effects, it is recommendable to select the appropriate operation strategy based on the real operating conditions.

In summary, Dr. Wenting Hou and Professor Hua Wei reported a data-driven robust unit commitment approach based on RKDE for coordination and optimization of thermal, hydro, wind, PV and nuclear power. The inclusion of peak load regulation helped improve the flexibility of NPPs. On the other hand, the water spillage strategies proved effective for balancing the interests of the basin and individual hydropower stations. This approach is superior to stochastic and conventional methods and can enhance the operation of power sources. In a statement to Advances in Engineering, the authors said that the study would enhance the stable operation of power systems.

Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/ photovoltaic/nuclear power systems - Advances in Engineering

About the author

Dr. Wenting Hou received his Ph.D. degree at Guangxi Key Laboratory of Power System Optimization and Energy Technology, Guangxi University in 2019. Currently, he is working with School of Mechanical and Electrical Engineering, Zhoukou Normal University. His current research interests include unit commitment, data-driven optimization and power system economic dispatch, etc.

.

About the author

Dr. Hua Wei He received the B.S. and M.S. degrees in power engineering from Guangxi University, Guangxi, China in 1981 and 1987, respectively, and Ph.D. degree in power engineering from Hiroshima University, Japan in 2002. He was a visiting Professor at Hiroshima University, Japan from 1994 to 1997. From 2004 to 2014, he was the vice-present of the Guangxi University. Now, he is a professor of Guangxi University. He is also the director of the Institute of Power System Optimization, Guangxi University. His research interests include power system operation and planning, particularly in the application of optimization theory and methods to power systems.

Reference

Hou, W., & Wei, H. (2021). Data-driven robust day-ahead unit commitment model for hydro/thermal/wind/photovoltaic/nuclear power systemsInternational Journal of Electrical Power & Energy Systems, 125, 106427.

Go To International Journal of Electrical Power & Energy Systems

Check Also

New Mechanism of Buoyant Plume Transition Unveiled by Large Eddy Simulations - Advances in Engineering

New Mechanism of Buoyant Plume Transition Unveiled by Large Eddy Simulations