Browsing by Author "Owusu, Frank Kofi"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Dynamic linear state space model for forecasting peak and short-term electricity demand using kalman filtered monte carlo method
    (August, 2018) Owusu, Frank Kofi; ; ;
    Electricity has become a major part of human life, especially in our part of the world. It is one of the most used energy across the world. Due to the fast changing world, the demand for electricity keeps on increasing from time to time yet there is not any efficient way of storing this energy for future use. So operators are very cautious about the amount to release and also to meet the demand of their consumers. For this reason, load forecasting has become a main integrated section in energy manage ment and production. This research seeks to look at Short-Term Load Forecasting. The objective is to forecast the peak demand and total energy generated or elec tricity demand. So the Seemingly Unrelated Time Series Equations Model which models the level or state and trend in the system was used for the study. A Markov Chain Monte Carlo (MCMC) method, Gibbs Sampler, together with the Kalman Filter and Kalman Smoother, the Forward Filtering Backward Sampling with Gibbs Sampler Algorithm were applied to the model to simulate for the variances also to predict the peak demand the next day’s peak and electricity demand. The running ergodic mean showed the convergence of the MCMC process and hence the posterior means of the variances were estimated. The one-step-ahead forecast showed a Mean Absolute Percentage error (MAPE) of 3.696% error in the peak demand forecast and a 4.235% error in the electricity demand forecast. The forecast for the next day was about 2187MW and 44090MW for the peak and electricity demands respectively. For further studies, the model can be extended to capture seasonal components.
  • No Thumbnail Available
    Item
    Seemingly unrelated time series model for forecasting the peak and short-term electricity demand: Evidence from the Kalman filtered Monte Carlo method
    (Heliyon, 2023-08) MARTIN, HENRY; Owusu, Frank Kofi; Amoako-Yirenkyi, Peter; Frempong, Nana Kena; Omari-Sasu, Akoto Yaw; Mensah, Isaac Adjei; Sakyi, Adu; 0000-0003-0173-1238
    In this extant paper, a multivariate time series model using the seemingly unrelated times series equation (SUTSE) framework is proposed to forecast the peak and short-term electricity demand using time series data from February 2, 2014, to August 2, 2018. Further the Markov Chain Monte Carlo (MCMC) method, Gibbs Sampler, together with the Kalman Filter were applied to the SUTSE model to simulate the variances to predict the next day’s peak and electricity demand. Relying on the study results, the running ergodic mean showed the convergence of the MCMC process. Before forecasting the peak and short-term electricity demand, a week’s prediction from the 28th to the 2nd of August of 2018 was analyzed and it found that there is a possible decrease in the daily energy over time. Further, the forecast for the next day (August 3, 2018) was about 2187 MW and 44090 MWh for the peak and electricity demands respectively. Finally, the robustness of the SUTSE model was assessed in comparison to the SUTSE model without MCMC. Evidently, SUTSE with the MCMC method had recorded an accuracy of about 96% and 95.8% for Peak demand and daily energy respectively
  • Thumbnail Image
    Item
    Time series arima modelling of inflation in Ghana: (1990 – 2009)
    (2010-08-14) Owusu, Frank Kofi
    Throughout the world, most central bank policy initiatives have been aimed at achieving and maintaining price stability and the Bank of Ghana is no exception to this rule. This study attempts to outline the practical steps which need to be undertaken in order to use the autoregressive integrated moving average (ARIMA) model for forecasting Ghana’s inflation. The main focus of the study is to model inflation and hence used to forecast the monthly inflation on short-term basis, for this purpose, different ARIMA models are used and the candid model is selected based on various diagnostic, evaluation and selection criteria. It can be concluded that the model has sufficient predictive powers and the findings are well in line with those of other studies. Again the study models inflation for the periods of 1990 to 2000 and 2001 to 2009 and it was realized that the inflation model for the period of 1990 to 2000 is ARIMA (1, 2,2) written as (y_t ) ̂=18.5770+0.455848t-3.57e^(-0.3) t^2+0.7807y_(t-1)-1.0813ε_(t-1)+0.1020ε_(t-2)+(ε_t ) ̂. Whilst that of 2001 to 2009 is modelled as ARIMA (2, 2, 1), written as (y_t ) ̂=34.3958-0.637228t+4.40e^(-0.3) t^2-1.3764y_(t-1)-0.4389y_(t-2)+0.9860ε_(t-1)+(ε_t ) ̂ It was concluded that inflation for the period of January 2001 to December 2009 was less than that of January 1990 to December 2000. The model is recommended for use by stakeholders because it has a lower error variance of ±1 which follows closely with the actual data. It is recommended further to be used as the basis for constructing deterministic models such as first and second order differential equations by future researchers