Monthly Load Forecasting Based on Optimum Grey Model

2011 ◽  
Vol 230-232 ◽  
pp. 1226-1230
Author(s):  
Ting Wang ◽  
Xi Miao Jia
Keyword(s):  
Historical Data ◽  
Load Forecasting ◽  
Forecast Accuracy ◽  
Forecast Model ◽  
Difficult Problem ◽  
Short Term ◽  
Original Series ◽  
Short Term Load Forecasting ◽  
Calculation Results ◽  
Long Time

Due to the variety and the randomicity of its influencing factors, the monthly load forecasting is a difficult problem for a long time. In order to improve the forecast accuracy, the paper proposes a new load forecast model based on improved GM (1, 1).First, the GM (1, 1) is used to forecast the load data, which takes the longitude historical data as original series, the increment trend of load was forecasted and takes the crosswise historical data as original series, the fluctuation trend of load was forecasted. On this basis the optimum method is led in. An optimal integrated forecasting model is built up. The case calculation results show that the proposed method can remarkably improve the accuracy of monthly load forecasting, and decrease the error. The integrated model this paper describes for short-term load forecasting is available and accurate.

scholarly journals A Two-Stage Short-Term Load Forecasting Method Using Long Short-Term Memory and Multilayer Perceptron

Energies ◽  
2021 ◽  
Vol 14 (18) ◽  
pp. 5873
Author(s):  
Yuhong Xie ◽  
Yuzuru Ueda ◽  
Masakazu Sugiyama
Keyword(s):  
Neural Networks ◽  
Multilayer Perceptron ◽  
Short Term Memory ◽  
Historical Data ◽  
Load Forecasting ◽  
Short Term ◽  
Term Memory ◽  
Proposed Model ◽  
Short Term Load Forecasting ◽  
Long Short Term Memory

Load forecasting is an essential task in the operation management of a power system. Electric power companies utilize short-term load forecasting (STLF) technology to make reasonable power generation plans. A forecasting model with low prediction errors helps reduce operating costs and risks for the operators. In recent years, machine learning has become one of the most popular technologies for load forecasting. In this paper, a two-stage STLF model based on long short-term memory (LSTM) and multilayer perceptron (MLP), which improves the forecasting accuracy over the entire time horizon, is proposed. In the first stage, a sequence-to-sequence (seq2seq) architecture, which can handle a multi-sequence of input to extract more features of historical data than that of single sequence, is used to make multistep predictions. In the second stage, the MLP is used for residual modification by perceiving other information that the LSTM cannot. To construct the model, we collected the electrical load, calendar, and meteorological records of Kanto region in Japan for four years. Unlike other LSTM-based hybrid architectures, the proposed model uses two independent neural networks instead of making the neural network deeper by concatenating a series of LSTM cells and convolutional neural networks (CNNs). Therefore, the proposed model is easy to be trained and more interpretable. The seq2seq module performs well in the first few hours of the predictions. The MLP inherits the advantage of the seq2seq module and improves the results by feeding artificially selected features both from historical data and information of the target day. Compared to the LSTM-AM model and single MLP model, the mean absolute percentage error (MAPE) of the proposed model decreases from 2.82% and 2.65% to 2%, respectively. The results demonstrate that the MLP helps improve the prediction accuracy of seq2seq module and the proposed model achieves better performance than other popular models. In addition, this paper also reveals the reason why the MLP achieves the improvement.

Short-Term Load Forecasting Using Hybrid Neural Network

2021 ◽  
Vol 12 (1) ◽  
pp. 142-156
Author(s):  
Muhammad Nadeem ◽  
Muhammad Altaf ◽  
Ayaz Ahmad
Keyword(s):  
Neural Network ◽  
New England ◽  
Low Cost ◽  
Electrical Power ◽  
Load Forecasting ◽  
Electricity Consumption ◽  
Forecast Model ◽  
Short Term ◽  
Short Term Load Forecasting ◽  
Multiple Variables

One of the important factors in generating low cost electrical power is the accurate forecasting of electricity consumption called load forecasting. The major objective of the load forecasting is to trim down the error between actual load and forecasted load. Due to the nonlinear nature of load forecasting and its dependency on multiple variables, the traditional forecasting methods are normally outperformed by artificial intelligence techniques. In this research paper, a robust short term load forecasting technique for one to seven days ahead is introduced based on particle swarm optimization (PSO) and Levenberg Marquardt (LM) neural network forecast model, where the PSO and LM algorithm are used for the training process of neural network. The proposed methods are tested to predict the load of the New England Power Pool region's grid and compared with the existing techniques using mean absolute percentage errors to analyze the performance of the proposed methods. Forecast results confirm that the proposed LM and PSO-based neural network schemes outperformed the existing techniques.

Middle – Long Electric Power Load Forecasting Based on GM(1,1) and Support Vector Machine

2010 ◽  
Vol 44-47 ◽  
pp. 2983-2987
Author(s):  
Xi Miao Jia ◽  
Guo Ping Song ◽  
Ting Wang ◽  
Feng Kong
Keyword(s):  
Support Vector Machine ◽  
Demand Forecasting ◽  
Forecast Accuracy ◽  
Forecast Model ◽  
Difficult Problem ◽  
Support Vector ◽  
Load Forecast ◽  
Power Load ◽  
Long Time ◽  
Load Vector

Due to the variety and the randomicity of its influencing factors, the electricity demand forecasting is a difficult problem for a long time. In order to improve the forecast accuracy, the paper proposes a new load forecast model based on GM(1,1) and support vector machine. First, the GM(1,1) is used to forecast the load data in the model. And then according to factors and historical load vector, support vector machine load forecast model is established to forecast the residuals of GM(1,1) and modified the forecast results of GM(1,1). Case analysis shows that the forecast method is suitable and effective, improving prediction precision compared with GM(1,1) and support vector machine, and has better utility value in mid-log term load forecast.

scholarly journals A Non-Linear Controller for Forecasting the Rising Demand for Electric Vehicles Applicable to Indian Road Conditions

2016 ◽  
Vol 6 (5) ◽  
pp. 2274
Author(s):  
Poorani S ◽  
Murugan R
Keyword(s):  
Neural Network ◽  
Historical Data ◽  
Load Forecasting ◽  
Weather Conditions ◽  
Ann Model ◽  
Short Term ◽  
The Neural Network ◽  
Short Term Load Forecasting ◽  
Temperature And Humidity ◽  
Charging Stations

<p class="Abstract">These days load forecasting is much more required  in order to reduce the wastage of energy. This paper is to implement &amp; develop the idea of short term load forecasting by using Artificial Neural Network, the design of the neural network model, input data selection and Training &amp; Testing by using short term load forecasting will be described in paper. For the EV load forecasting only 2 variables are being used as temperature and humidity to forecast the output as load. This type of designed ANN model will be mapped by using historical data of temperature and humidity (taken from meteorological sites), whereas it is being Trained &amp; Tested by using historical data of loading of EV charging stations (Chetan maini ,Bangalore) of a particular area as Coimbatore to give the desired result. Training &amp; Testing done by using large amount of historical data of weather conditions and loading data (kV). By the help of this model they can predict their daily loads (next day's load) by putting historical data in the acquired algorithm.</p>

Based on the EMD and PSO-BP Neural Network of Short-Term Load Forecasting

2012 ◽  
Vol 614-615 ◽  
pp. 1872-1875 ◽  
Author(s):  
Feng Sha ◽  
Feng Zhu ◽  
Shun Nan Guo ◽  
Jian Tong Gao
Keyword(s):  
Neural Network ◽  
Bp Neural Network ◽  
Load Forecasting ◽  
Intrinsic Mode Functions ◽  
Short Term ◽  
Network Training ◽  
Short Term Load Forecasting ◽  
Long Time ◽  
Neural Network Algorithm ◽  
Bp Model

This paper proposes that based on the EMD and PSO-BP neural network of short-term load forecasting. This method will be automatically historical load sequence into several independent intrinsic mode functions (IMF) by using EMD. As the BP neural network training for a long time and easy to fall into local minimum of the shortcomings, we use genetic algorithm to optimize BP neural network algorithm replaces the traditional BP algorithm. Finally, we use the BP neural network optimized separately for each IMF component for training and prediction. We should add the component to the final prediction forecast. This method has higher precision than the EMD-BP model prediction by Simulation results show.

A Non-Linear Controller for Forecasting the Rising Demand for Electric Vehicles Applicable to Indian Road Conditions

2016 ◽  
Vol 6 (5) ◽  
pp. 2274
Author(s):  
Poorani S ◽  
Murugan R
Keyword(s):  
Neural Network ◽  
Historical Data ◽  
Load Forecasting ◽  
Weather Conditions ◽  
Ann Model ◽  
Short Term ◽  
The Neural Network ◽  
Short Term Load Forecasting ◽  
Temperature And Humidity ◽  
Charging Stations

<p class="Abstract">These days load forecasting is much more required  in order to reduce the wastage of energy. This paper is to implement &amp; develop the idea of short term load forecasting by using Artificial Neural Network, the design of the neural network model, input data selection and Training &amp; Testing by using short term load forecasting will be described in paper. For the EV load forecasting only 2 variables are being used as temperature and humidity to forecast the output as load. This type of designed ANN model will be mapped by using historical data of temperature and humidity (taken from meteorological sites), whereas it is being Trained &amp; Tested by using historical data of loading of EV charging stations (Chetan maini ,Bangalore) of a particular area as Coimbatore to give the desired result. Training &amp; Testing done by using large amount of historical data of weather conditions and loading data (kV). By the help of this model they can predict their daily loads (next day's load) by putting historical data in the acquired algorithm.</p>

One-Hour Ahead Load Forecasting Based on Wavelet Neural Networks

2015 ◽  
Vol 785 ◽  
pp. 53-57 ◽  
Author(s):  
Narin Sovann ◽  
Perumal Nallagownden ◽  
Zuhairi Baharudin
Keyword(s):  
Neural Network ◽  
New England ◽  
Wavelet Transforms ◽  
Load Forecasting ◽  
Forecast Model ◽  
Wavelet Neural Networks ◽  
Short Term ◽  
Load Forecast ◽  
Proposed Model ◽  
Short Term Load Forecasting

This paper presents the improvement on accuracy and reliability of the load forecast model. It is well-known that characteristics of a load series is a non-stationary data, which is a constraint for the load forecast methods to achieve accurate and robustness responses. To overcome this limitation, a synergized method between wavelet transform and artificial neural network is proposed for short-term load forecasting. The modeling processes such as minimizing distorted data due to convolution operator of the wavelet transforms, model inputs and neural network design are presented. The proposed method is tested using historical load data of independent system operation New England. The results of the proposed model significantly outperform either accuracy or robustness results over neural network model.

Sign in / Sign up

Export Citation Format

Share Document