Litcius/Paper detail

Residential Load Time Series Forecasting using ANN and Classical Methods

Lekshmi R. Chandran, Nikhil Jayagopal, Lekshmi Shyma Lal, Chaithanya Narayanan, S. Deepak, V. Harikrishnan

202120 citationsDOI

Abstract

Electrical energy consumption is affected by variations in temperature, seasons, humidity, weekdays and festive occasions. It affects electric load forecasting. This results in irregularity of production and distribution. Residential load forecasting can effectively help individual consumers to regulate their energy consumption and reduce cost. This in effect help utility to regulate the power consumed. This paper focuses on forecasting techniques for residential load. This paper compares the traditional load forecasting method with the neural network. The traditional forecasting models studied are ARIMA and ARIMAX. In the Neural Network, different models using feedforward backpropagation networks are studied. It is found that artificial neural networks give better results compared to traditional time series forecasting methods. The superiority of artificial neural network in enhancing load forecasting and the effect of transfer function used in the neural network is discussed in this paper.

Topics & Concepts

Artificial neural networkAutoregressive integrated moving averageComputer scienceBackpropagationFeedforward neural networkTime seriesElectrical loadDemand forecastingEnergy consumptionElectric power systemFeed forwardArtificial intelligenceSeries (stratigraphy)Machine learningPower (physics)EngineeringOperations researchControl engineeringVoltageElectrical engineeringPaleontologyQuantum mechanicsPhysicsBiologyEnergy Load and Power ForecastingImage and Signal Denoising MethodsBuilding Energy and Comfort Optimization
Residential Load Time Series Forecasting using ANN and Classical Methods | Litcius