Spatial Electric Load Forecasting

Electricity load forecasting is always a key instrument for the effective operation and planning of power systems. This paper presents our recent works on short-term electricity demand forecasting for an electric utility in Midwest US focusing on day-ahead operation and market. The target system covers a large geographical area, and several alternative meteorological forecasts are available from different commercial weather services.For a system with large geographical area, a single model for load forecasting of the entire area sometimes cannot guarantee satisfactory forecasting accuracy because of the load diversity. We therefore develop a multi-region load forecasting model, which can find the optimal region partition under diverse weather and load conditions and finally achieve more accurate forecasts for aggregated system demand. Furthermore, to effectively take advantage of the alternative meteorological predictions in the load forecasting system, combining forecasting using adaptive coefficients is applied to share the strength of the different temperature forecasts. The proposed forecasting system has been tested by using the real data from the system. A range of comparisons with different forecasting models have been conducted. The forecasting results demonstrate the superiority of the proposed methodology. Shu Fan (M'08-SM'09) received the B.S., M.S., and Ph.D. degrees in the

show abstract

“…The level of diversity for a group of electrical loads has been defined by a coincidence factor C [6] [7],…”

Section: Load Diversity Analysismentioning

confidence: 99%

Multi-region load forecasting considering alternative meteorological predictions

Fan

Methaprayoon²,

Lee

2010

IEEE PES General Meeting

View full text Add to dashboard Cite

show abstract

“…The magnitude of σ ε 2 can be derived using a peak load coincidence curve, which is a commonly used metric to represent the ratio of the aggregate peak load to the individual peaks of M customers [16]. If the load is modeled as a normally distributed random variable, the "peak" of a single or group of customers can be defined as the mean plus some multiple of the standard deviation.…”

Section: B Load Modelmentioning

confidence: 99%

Reliability evaluation of islanded microgrids with stochastic distributed generation

Kennedy¹

2009

2009 IEEE Power &Amp; Energy Society General Meeting

View full text Add to dashboard Cite

Microgrids can potentially improve distribution system reliability when enabled to operate in an islanded mode. Future microgrids will likely contain substantial amounts of stochastic generation, which have a more uncertain impact on reliability as compared to dispatchable generation sources. New methods that realistically represent the stochastic behavior of these sources and their impact on reliability need to be developed. This paper proposes a new methodology for assessing local generation adequacy for an islanded microgrid with limited stochastic resources. Instead of modeling the generation and load independently, a combined generation-to-load ratio (GLR) is modeled as a Markov process, thereby preserving the correlation among disperse photovoltaic installations and the load. Historical data on solar irradiation and electric load are scaled and smoothed to account for their spatial distribution and used to estimate the Markov transition rates and the final reliability indices. The process illustrates the impact of stochastic behavior on reliability in a transparent manner and is useful for evaluating the reliability benefits of load curtailment and increasing local generation capacity.

show abstract

“…Some have simply noted that no such test is available that incorporates the spatial correlation across predictions (Longhi and Nijkamp 2007). Many give point estimates of predictive accuracy or choose the model that minimizes some loss function, but they may not quantify the uncertainty associated with those estimates or include potential spatial dependence in their estimates; for example, see the works of Atger (2003), Gong, Barnston, and Ward (2003), Kleiber et al (2011), Willis (2002).…”

Section: Introductionmentioning

confidence: 98%

“…Currently, predicting wind speeds for wind power generation is a particularly important area of application in which such a prediction comparison test would be beneficial (Willis 2002;Genton and Hering 2007). No cost-effective method for storing wind energy exists, so it must be used as soon as it is produced.…”

Section: Introductionmentioning

confidence: 99%

Comparing Spatial Predictions

Hering

Genton

2011

Technometrics

View full text Add to dashboard Cite

Under a general loss function, we develop a hypothesis test to determine whether a significant difference in the spatial predictions produced by two competing models exists on average across the entire spatial domain of interest. The null hypothesis is that of no difference, and a spatial loss differential is created based on the observed data, the two sets of predictions, and the loss function chosen by the researcher. The test assumes only isotropy and short-range spatial dependence of the loss differential but does allow it to be non-Gaussian, non-zero-mean, and spatially correlated. Constant and nonconstant spatial trends in the loss differential are treated in two separate cases. Monte Carlo simulations illustrate the size and power properties of this test, and an example based on daily average wind speeds in Oklahoma is used for illustration. Supplemental results are available online.

show abstract

Spatial Electric Load Forecasting

Cited by 97 publications

References 0 publications

Multi-region load forecasting considering alternative meteorological predictions

Multi-region load forecasting considering alternative meteorological predictions

Reliability evaluation of islanded microgrids with stochastic distributed generation

Comparing Spatial Predictions

Contact Info

Product

Resources

About