Mid-Long Term Electricity Consumption Forecasting Analysis Based on Cyber-Physical-Social System Architecture

Abstract

Daily electricity consumption forecasting is a classical problem. Existing forecasting algorithms tend to have decreased accuracy on special dates like holidays. This study decomposes the daily electricity consumption series into three components: trend, seasonal, and residual, and constructs a two-stage prediction method using piecewise linear regression as a filter and Dilated Causal CNN as a predictor. The specific steps involve setting breakpoints on the time axis and fitting the piecewise linear regression model with one-hot encoded information such as month, weekday, and holidays. For the challenging prediction of the Spring Festival, distance is introduced as a variable using a third-degree polynomial form in the model. The residual sequence obtained in the previous step is modeled using Dilated Causal CNN, and the final prediction of daily electricity consumption is the sum of the two-stage predictions. Experimental results demonstrate that this method achieves higher accuracy compared to existing approaches.

Optimized Gated Recurrent Unit for Mid-Term Electricity Price Forecasting

“…7. It can be summarized that the proposed forecasting model for Australian electricity market is feasible with MAPE value in the range of 0.17% to 1.68% as compared to previous works with MAPE values of approximately 11% in [17], 9% in [2], 3% to 5% in [18], 1.9% in [19], 5.85% to 11.8% in [28], and 2.4% to 4.3% in [15]. The MAPE value was computed by averaging the MAPE values for the five states of Australia that are focused in this work.…”

“…ARIMA [18] DBN [18] Mid-long term electricity consumption wuhan, china 5.140% 3.278% Data analysis is limited since short-term prediction is challenging.…”

“…Nowadays, researchers are capable of employing and extracting complex information from time series data to solve various problems such as wind speed prediction [4], stock market prediction [3,[5][6][7], and forecasting electricity prices [8,9]. Time series models have been widely used to forecast electricity prices, although they can be challenging due to large variations [10][11][12][13][14][15][16][17][18][19]. Existing statistical techniques tried to reveal the specific pattern of historic power price utilizing curve fitting.…”

“…The advantages of this method include good precision, model functionality, and reliability. Meanwhile, Autoregressive Integrated Moving Average (ARIMA) was proposed for electricity and power load forecasting [13,14,16,18]. However, application of statistical models had shown to be challenging when predicting multi-dimensional nonlinear price of electricity since they are mainly based on linear equations.…”

“…Recurrent neural network (RNN) which is a deep learning technique with a recurrent feedback network has shown to perform better when dealing with time series data compared to conventional machine learning algorithm [28]. The work in [18] proposed a Deep Belief Network (DBN) for electricity price forecasting. On the other hand, the work in [15], proposed LSTM network for electricity price forecasting.…”