A transfer learning-based scenario generation method for stochastic optimal scheduling of microgrid with newly-built wind farm

Energy storage system (ESS) can play a positive role in the power system due to its ability to store, charge and discharge energy. Additionally, it can be installed in various capacities, so it can be used in the transmission and distribution system and even at home. In this paper, the proposed algorithm for economic optimal scheduling of ESS linked to transmission systems in the Korean electricity market is proposed and incorporated into the BESS (battery energy storage system) demonstration test center. The proposed algorithm considers the energy arbitrage operation through SMP (system marginal price) and operation considering the REC (renewable energy certification) weight of the connected wind farm and frequency regulation service. In addition, the proposed algorithm was developed so that the SOC (state-of-charge) of the ESS could be separated into two virtual SOCs to participate in different markets and generate revenue. The proposed algorithm was simulated and verified through Matlab and loaded into the demonstration system using the Matlab “Runtime” function.

Download Full-text

Reliability Assessment of the Wind-solar-battery Storage System Based on Optimized Configuration

ITM Web of Conferences ◽

10.1051/itmconf/20181701018 ◽

2018 ◽

Vol 17 ◽

pp. 01018

Author(s):

Yuan Wang ◽

Haojie Liu

Keyword(s):

Wind Farm ◽

Power Grid ◽

Storage System ◽

Thermal Power ◽

Clean Energy ◽

Optimal Scheduling ◽

Battery Storage ◽

Optimal Dispatch ◽

New Energy ◽

The Impact

In order to analyze the impact of new energy power generation on the power grid system, the reliability evaluation of the wind-solarbattery storage system is carried out. Proposed to wind power, solar, thermal power, different sodium-sulfur battery storage combined optimal dispatch of scenery. The shortest variance of the net load and the maximum variance of the wind storage system are taken as the objective function. The short-term optimal scheduling model of the power grid is established based on the characteristics of the wind farm, the characteristics of the solar field and the electric field of the sodium flow battery. Multi-objective particle swarm optimization The algorithm solves the model and obtains the output power of wind, light, storage and fire under different new energy strategies. The reliability is evaluated by Monte-Carlo method. Taking the IEEE-30 node as an example, it is proved that the proposed model is reasonable and the new energy can improve the clean energy consumption ability and minimize the impact on the power grid under the optimal scheduling strategy.

Download Full-text

Ultra-Short-Term Wind Power Prediction Based on Bidirectional Gated Recurrent Unit and Transfer Learning

Frontiers in Energy Research ◽

10.3389/fenrg.2021.808116 ◽

2021 ◽

Vol 9 ◽

Author(s):

Wenjin Chen ◽

Weiwen Qi ◽

Yu Li ◽

Jun Zhang ◽

Feng Zhu ◽

...

Keyword(s):

Wind Power ◽

Transfer Learning ◽

Wind Farm ◽

Meteorological Data ◽

Prediction Method ◽

Wind Farms ◽

Power Prediction ◽

Short Term ◽

Training Time ◽

Gated Recurrent Unit

Wind power forecasting (WPF) is imperative to the control and dispatch of the power grid. Firstly, an ultra-short-term prediction method based on multilayer bidirectional gated recurrent unit (Bi-GRU) and fully connected (FC) layer is proposed. The layers of Bi-GRU extract the temporal feature information of wind power and meteorological data, and the FC layer predicts wind power by changing dimensions to match the output vector. Furthermore, a transfer learning (TL) strategy is utilized to establish the prediction model of a target wind farm with fewer data and less training time based on the source wind farm. The proposed method is validated on two wind farms located in China and the results prove its superior prediction performance compared with other approaches.

Download Full-text

Using Transfer Learning to Build Physics-Informed Machine Learning Models for Improved Wind Farm Monitoring

Energies ◽

10.3390/en15020558 ◽

2022 ◽

Vol 15 (2) ◽

pp. 558

Author(s):

Laura Schröder ◽

Nikolay Krasimirov Dimitrov ◽

David Robert Verelst ◽

John Aasted Sørensen

Keyword(s):

Transfer Learning ◽

Wind Farm ◽

Prediction Performance ◽

Data Driven ◽

Monitoring Systems ◽

Blade Angle ◽

Behavior Monitoring ◽

Normal Behavior ◽

Artificial Neural Network Ann ◽

Machine Learning Models

This paper introduces a novel, transfer-learning-based approach to include physics into data-driven normal behavior monitoring models which are used for detecting turbine anomalies. For this purpose, a normal behavior model is pretrained on a large simulation database and is recalibrated on the available SCADA data via transfer learning. For two methods, a feed-forward artificial neural network (ANN) and an autoencoder, it is investigated under which conditions it can be helpful to include simulations into SCADA-based monitoring systems. The results show that when only one month of SCADA data is available, both the prediction accuracy as well as the prediction robustness of an ANN are significantly improved by adding physics constraints from a pretrained model. As the autoencoder reconstructs the power from itself, it is already able to accurately model the normal behavior power. Therefore, including simulations into the model does not improve its prediction performance and robustness significantly. The validation of the physics-informed ANN on one month of raw SCADA data shows that it is able to successfully detect a recorded blade angle anomaly with an improved precision due to fewer false positives compared to its purely SCADA data-based counterpart.

Download Full-text

An economic and low-carbon day-ahead Pareto-optimal scheduling for wind farm integrated power systems with demand response

ENERGYO ◽

10.1515/energyo.s40565-014-0094-7 ◽

2018 ◽

Author(s):

Rui MA ◽

Kai LI ◽

Xuan LI ◽

Zeyu QIN

Keyword(s):

Power Systems ◽

Demand Response ◽

Wind Farm ◽

Optimal Scheduling ◽

Pareto Optimal ◽

Low Carbon ◽

Integrated Power Systems

Download Full-text

Optimal Scheduling of Microgrid Based on Deep Deterministic Policy Gradient and Transfer Learning

Energies ◽

10.3390/en14030584 ◽

2021 ◽

Vol 14 (3) ◽

pp. 584

Author(s):

Luqin Fan ◽

Jing Zhang ◽

Yu He ◽

Ying Liu ◽

Tao Hu ◽

...

Keyword(s):

Transfer Learning ◽

Supply And Demand ◽

Optimization Methods ◽

Optimal Scheduling ◽

Action Space ◽

Operation Mechanism ◽

Scheduling Strategy ◽

Complex Operation ◽

Policy Gradient ◽

Discrete Action

Microgrid has flexible composition, a complex operation mechanism, and a large amount of data while operating. However, optimization methods of microgrid scheduling do not effectively accumulate and utilize the scheduling knowledge at present. This paper puts forward a microgrid optimal scheduling method based on Deep Deterministic Policy Gradient (DDPG) and Transfer Learning (TL). This method uses Reinforcement Learning (RL) to learn the scheduling strategy and accumulates the corresponding scheduling knowledge. Meanwhile, the DDPG model is introduced to extend the microgrid scheduling strategy action from the discrete action space to the continuous action space. On this basis, this paper holds that a microgrid optimal scheduling TL algorithm on the strength of the actual supply and demand similarity is proposed with a purpose of making use of the existing scheduling knowledge effectively. The simulation results indicate that this paper can provide optimal scheduling strategy for microgrid with complex operation mechanism flexibly and efficiently through the effective accumulation of scheduling knowledge and the utilization of scheduling knowledge through TL.

Download Full-text