Artificial Neural Network and Particle Swarm Optimization for Medium Term Electrical Load Forecasting in a Smart Campus

Energy demand has continued to increase rapidly not exempting Covenant University. As the university continues to witness infrastructural expansion and population increase, it has become a necessity for energy consumption to be predicted. Hence, this research work developed a medium-term load forecasting system to solve this problem and ensure an efficient electricity supply from the power system operators of Covenant University. The forecast was carried out on real-time monthly load data collected from the university community power plant between 2015 and 2018, using the Artificial Neural Network (ANN) model. A medium-term load forecast was evaluated based on three different ANN algorithms. The FeedForwardNet, Cascadeforwardnet and Fitnet are tested against three (3) different learning algorithms namely Levenberg Marquardt, Bayesian regularization and BFGS quasi-Newton backpropagation with a particle swarm optimizer. And the network performance was obtained using Normalized Root Mean Square Error (nRMSE %). The result revealed an nRMSE of 0.0634%, a correlation coefficient, r, of 0.9082 and the fastest computation speed of 171.789 seconds. Hence, this study provides a point of reference for other related studies and future energy forecast improvement in the study location.

[1]  M. Raza,et al.  On recent advances in PV output power forecast , 2016 .

[2]  K. El-Naggar,et al.  HEURISTIC APPROACH FOR OPTIMAL PARAMETER ESTIMATION OF ELECTRIC LOAD FORECAST MODEL , 2009 .

[3]  K. M. El-Naggar,et al.  Electric Load Forecasting Using Genetic Based Algorithm, Optimal Filter Estimator and Least Error Squares Technique: Comparative Study , 2007 .

[4]  H. K. Temraz,et al.  Applications of noniterative least absolute value estimation for forecasting annual peak electric power demand , 1998, Canadian Journal of Electrical and Computer Engineering.

[5]  Xi Jin,et al.  Medium and Long-Term Load Forecast Based on PSO-LS-SVR , 2012 .

[6]  S. Hyakin,et al.  Neural Networks: A Comprehensive Foundation , 1994 .

[7]  Yuhong Li,et al.  Parameter optimization of nonlinear grey Bernoulli model using particle swarm optimization , 2009, Appl. Math. Comput..

[8]  A. A. Awelewa,et al.  ARTIFICIAL NEURAL NETWORK BASE SHORT-TERM ELECTRICITY LOAD FORECASTING: A CASE STUDY OF A 132/33KV TRANSMISSION SUB-STATION , 2020 .

[9]  Idakwo O. Harrison,et al.  Short Term Electric Load Forecasting of 132/33KV Maiduguri Transmission Substation using Adaptive Neuro-Fuzzy Inference System (ANFIS) , 2014 .

[10]  Tao Hong,et al.  Short Term Electric Load Forecasting , 2012 .

[11]  Niu Dongxiao,et al.  Research of Residual Error-Ant Colony Optimization Gray Model Based on Markov in Load Forecasting , 2008, 2008 ISECS International Colloquium on Computing, Communication, Control, and Management.

[12]  I. A. Odigwe,et al.  A Comparative Study of Regression Analysis and Artificial Neural Network Methods for Medium-Term Load Forecasting , 2017 .

[13]  B. Tech,et al.  Application of Neural Network to Load Forecasting in Nigerian Electrical Power System , 2007 .

[14]  A. Adewale,et al.  Energy audit and optimal power supply for a commercial building in Nigeria , 2018 .

[15]  Nadeem Javaid,et al.  Short-Term Load Forecasting in Smart Grids: An Intelligent Modular Approach , 2019, Energies.

[16]  A. A. Awelewa,et al.  Medium-Term Load Forecasting Of Covenant University Using The Regression Analysis Methods , 2014 .

[17]  Mahrufat D. Olagoke.  Short Term Electric Load Forecasting using Neural Network and Genetic Algorithm , 2016 .