Towards Developing a Decision Support System for Electricity Load Forecast

Abstract

Short-term load forecasting STLF is an essential procedure for effective and efficient realtime operations planning and control of generation within a power system. It provides the basis for unit-commitment and power system planning procedures, maintenance scheduling, system security assessment, and trading schedules. It establishes the generation, capacity, and spinning reserve schedules which are posted to the market. Without optimal load forecasts, additional expenses due to uneconomic dispatch, over/under purchasing, and reliability uncertainty can cost a utility millions of dollars [ ].Many approaches have been considered for STLF. The benefits of increased computational power and data storage have enhanced the capabilities of artificial intelligence methods for data analysis within the power industry [ ]. Yet, even with the advancements of technology, industry forecasts are often based on a traditional similar day forecasting methodology or rigid statistical models with reduced variable modifiers for forecasting aggregated system load.Power systems with a large spatial presence provide an increased challenge for load forecasters, as they often face large diversity within their load centres as well as diverse weather conditions. These geographically separated load centres often behave independently and add considerable complexity to the system dynamics and forecasting procedure.This chapter investigates forecasting of electrical demand at an electric utility within the province of Saskatchewan, Canada and proposes a multi-region load forecasting system based on weather-related demand variables. The control area examined consists of over , kilometres of power lines with transmission voltages of , , and KV. The control area was apportioned into twelve load centres, consisting primarily of conforming loads. These conforming loads were cities and rural load clusters not including large indus-

Daily Natural Gas Load Forecasting Based on a Hybrid Deep Learning Model

“…Actually, natural gas load is affected by various factors. In daily NGLF, weather factors are widely used because they are easily accessible and highly relevant [6,43]. In this paper, 14 weather factors were selected, including the lowest temperature (LT), average temperature (AT), highest temperature (HT), lowest dew point (LD), average dew point (AD), highest dew point (HD), lowest humidity (LH), average humidity (AH), highest humidity (HH), lowest visibility (LV), average visibility (AV), average air pressure (AA), average wind speed (AW), and precipitation (PP).…”

A Taxonomy of electricity demand forecasting techniques and a selection strategy

“…This technique can also be combined with artificial neural networks for reducing forecasting errors. This method may be rated as of low cost and moderately complex suitable for shortterm load forecasts with medium accuracy level and high skills using knowledge discovery in databases, and high accuracy level and medium skills using artificial neural networks [40][43] [44]. The method requires large data sets [40] [44].…”

“…Normally, temperature is exponentially smoothened before including in the model (the exponential smoothing technique is described in the next section) [41]. Exponential smoothing may be essential to run preprocessing commands for normalizing the loads that may have increased or decreased following an unusual trend because of unusual weather conditions [40].…”

“…In addition, the model may not be able to predict interrelationships between two or more seasonal fluctuations [47]. The model is useful for short-term electricity load forecasting with moderate accuracy [40] [47]. However, it requires loads of data and high level of expertise for ensuring reliable forecasting outputs.…”

A Taxonomy of electricity demand forecasting techniques and a selection strategy

“…However, it requires loads of data and high level of expertise for ensuring reliable forecasting outputs. The process for forecasting is presented in the Figure 2 [40]. Figure 2: Electricity load demand forecasting by combining exponential smoothing and multiple regression analysis [40] The data is normalized by eliminating unusual eventbased fluctuations (like, a major power cut due to a grid failure) and then exponential smoothing of seasonal fluctuations is applied [40].…”