A Simple and Reliable PV Forecasting Method for Local Area Energy Management

Sasaki, Yuya; Yorino, Naoto; Wahyudi, F. I.; Seikoba, Dai; Mitsumasa, Asada; Ma, Liying; Zoka, Yoshifumi

doi:10.1541/ieejpes.137.538

Cited by 4 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We first carry out PV output forecast and its prediction error analysis. A weather clustering method has been proposed for this purpose in [19]. A covariance matrix of PV output errors is calculated, while CIs of the PV forecast errors are specified depending on required reliability against system collapse.…”

Section: Proposed Approach To Treat Uncertaintiesmentioning

confidence: 99%

“…The UC can be fully utilized in the optimization process. Uncertainties related to the prediction and fluctuation of PV are handled particularly [19].…”

Section: Day-ahead Planning Managermentioning

confidence: 99%

“…The system consists of 30-generators, 31-load areas with RESs and a BT station in Table 1. PV G23 G28 G27 G26 G25 G24 G15 G14 G13 G12 G11 G29 G22 G21 G20 G19 G18 G17 G16 G10 G9 G8 G7 G6 G5 G4 G3 G2 G1 G30 1 Forecasting data are shown in Figure 4 in case of sunny day, which has been given by our forecasting method [19]. Typical load patterns are used in the proposed optimization process.…”

Section: Case Studies 41 Conditions Of the Simulationmentioning

confidence: 99%

See 2 more Smart Citations

Real-time dynamic economic load dispatch integrated with renewable energy curtailment

Sasaki¹,

Tsurumi²,

Yorino³

et al. 2019

Journal of International Council on Electrical Engineering

Self Cite

View full text Add to dashboard Cite

Recent power system operations are faced with the difficulties due to large penetration levels of intermitted renewable energy sources (RESs), for example, photovoltaic power (PV) generations and wind-turbine (WT) generations. Operations of a large number of uncontrollable RES also cause an unforeseen issue for the reliability in power system operation. To tackle the issue, this paper proposes a novel real-time generation schedule (GS) integrated with RES' power curtailment. The proposed real-time GS become to effectively compute the outputs of the associated generating units so as to meet the required electricity load. The GS computation is frequently updated every 5-min in real-time operation. The proposed approach is applied for real-time dynamic economic load dispatch for a real-time GS that completely treats the limited controllable energy resources under the uncertainties. Furthermore, real-time PV forecasting with their covariance matrices of the prediction errors is effectively utilized in the approach. In case of a unforeseen situation when the estimated electricity load cannot match the controllable unit's capability, the approach can be detect a minimum amount of supply-demand mismatch in advance, and can manage it reliably for the considerable time horizon.

show abstract

Section: Proposed Approach To Treat Uncertaintiesmentioning

confidence: 99%

“…The UC can be fully utilized in the optimization process. Uncertainties related to the prediction and fluctuation of PV are handled particularly [19].…”

Section: Day-ahead Planning Managermentioning

confidence: 99%

Section: Case Studies 41 Conditions Of the Simulationmentioning

confidence: 99%

See 1 more Smart Citation

Real-time dynamic economic load dispatch integrated with renewable energy curtailment

Sasaki¹,

Tsurumi²,

Yorino³

et al. 2019

Journal of International Council on Electrical Engineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…In recent years, the penetration of renewable energy has been promoted from the perspective of preventing global warming, and the massive introduction of photovoltaic power generation systems, mainly in residential areas, is expected in the future. There are many research papers that forecast the output of renewable energy sources, such as wind and solar power, a few minutes or hours later [1]- [3]. In these papers, prediction methods using neural network, which are part of AI, are proposed.…”

Section: Introductionmentioning

confidence: 99%

Effect of Demand Forecast Error on Imbalance Cost in Residential Micro Grid

Nozaki¹,

Kumano²

2022

REPQJ

View full text Add to dashboard Cite

In this paper, we conduct electric power simulation by forecasting electricity demand in general households with PV power generation based on the current electricity supply and demand system in Japan. Specifically, forecast errors for electricity demand in general households are taken into account, and machine learning is used to forecast the output of residential photovoltaic power generation. The actual electricity demand, which is the difference between the electricity demand in general households and the PV output, is then calculated, and the degree of contribution of the assumed model to the electricity system is evaluated using an index called imbalance cost from the perspective of a retail electricity provider.

show abstract

“…For the reasons described above, the studies on the forecast of PV output are actively investigated [5][6][7]. However, when forecasting output of PV, the forecasting results will be specialized in installation environment and characteristics of the panel.…”

Section: Introduction mentioning

confidence: 99%

Insolation Forecasting Method Using NN with EPSO in Consideration of Correlation between Other Regions

Shoji¹,

Hoga²

2018

JEE

View full text Add to dashboard Cite

Recently, there is a movement that aims to realize a distributed energy system in Japan. The PV (photovoltaic) generation is especially expected, and it is anticipated that interconnection number of PVs will increase more and more in the future. However, an amount of insolation is easily affected by the weather, the output of PV and the supply of electricity becomes unstable. And it causes various problems in a distribution system. Therefore, it is important to forecast the amount of insolation. In previous study, MLPNN (multilayer perceptron neural network) is a general method to forecast the amount of insolation. However, it is in danger of falling into a local solution by only MLPNN. In this study, the authors propose a forecasting method of amount of insolation using MLPNN and EPSO (evolutionary particle swarm optimization). The authors use EPSO in addition to MLPNN to solve the problem. The authors also propose a forecasting method of amount of insolation using other regions weather data for the accuracy improvement.

show abstract

A Simple and Reliable PV Forecasting Method for Local Area Energy Management

Cited by 4 publications

References 4 publications

Real-time dynamic economic load dispatch integrated with renewable energy curtailment

Real-time dynamic economic load dispatch integrated with renewable energy curtailment

Effect of Demand Forecast Error on Imbalance Cost in Residential Micro Grid

Insolation Forecasting Method Using NN with EPSO in Consideration of Correlation between Other Regions

Contact Info

Product

Resources

About