Now showing 1 - 5 of 41
- ItemOpen AccessRecurrent Neural Network Model for Forecasting Electricity Demand in Nigeria(Journal of Engineering Research, 2016-12) Abdulsalam, K.A.; Adegbenro, O; Akinbulire, T.O.This work uses modular recurrent neural network to estimate the electricity demand in Nigeria from 2015 to 2050. The network is a 2-layer multi-input, single-output model with twelve neurons trained using Levenberg-Marquardt algorithm. The data structure used for training is cell array of sequential concurrent data. The Recurrent Neural Network model was simulated as Non-linear Auto Regressive with eXogenous (NARX) model in Matlab environment and the predicted load for 2015 is about 550GWh; and an expected demand increase of 7.5 % every five year.
- ItemOpen AccessSelection of an electricity tariff plan for mini-grid business models: an intuitionistic fuzzy axiomatic design approach(Springer Nature Singapore Pte Ltd. 2020, 2020-02-06) Babatunde, O.M.; Ighravwe, D.E.; Oluseyi, P.O.; Mashao, D.This article proposes a conceptual framework that used mini-gird business models (MBMs) to rank electricity tariff plans. The framework combines intuitionistic fuzzy set (IFS), Criteria Importance Through Inter-criteria Correlation (CRITIC), fuzzy axiomatic design (FAD) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). It considered when decision-makers specify or did not specify for MBM’s requirements. Also, this article considered uncertainty in decision-makers information. The framework applicability was showed using information obtained from Sub-Sahara Africa. Five decision-makers provided information for the model implementation based on a well-structured questionnaire. The CRITIC method results showed that least and most important MBM are community-based and private MBMs, respectively. The IFS-FAD method identified time-based electricity tariff plan as the best suitable for the case study, while the IFS-TOPSIS method identified flexible best electricity tariff plans for the case study, respectively. However, both methods identified flat-rate as the least suitable plan. The proposed framework will provide mini-grid business owners with relevant information for energy tariff selection.
- ItemOpen AccessPower Quality Enhancement for a Standalone Photovoltaic System(Faculty of Engineering, University of Benin, Benin City, Nigeria, 2019-12) Oluseyi, P.O.; Akinbulire, P.O.; Ugherughe, J.; Oladoyinbo, O.; Babatunde, O.M.A distribution network connected to rural consumers is often quite weak due to the long distance. Hence, an increase in power demand over this network will result to power quality problems. The main cause of this unacceptable situation is the inability of the power system to meet the demand of reactive power. Inductive loads are the main cause of power quality problems and they are widely used in domestic and industrial sector, contributing to inefficient use of energy. In this paper, a reactive power compensator is proposed to give voltage stabilization, good power quality and to control non-linear loads in a standalone photovoltaic system. The results obtained show significant enhancement in terms of power factor correction of 0.87, thereby improving the power quality of the system.
- ItemOpen AccessComparative Load Flow Analysis of UNILAG Power Distribution Network using Newton Raphson and Gauss Seidel Methods(Faculty of Engineering, University of Benin, Benin City, Nigeria, 2019-06) Akinbulire, T.O.; Oluseyi, P.O.; Udoakam, G.A.; Babatunde, O.M.The evaluation of power flow in the distribution network has many techniques but there has been a very much interest in the traditionally known methods. These methods have enjoyed very wide acceptability and applicability. However, a comparative study of these techniques for the investigation of the load flow analysis for any of the Nigeria’s distribution systems is not adequately and proficiently documented. Thus, this inspired the adaptation of these techniques for the solution of a structured distribution network in the University of Lagos (UNILAG) Campus. The opportunity presented by this research is the deployment of these methods for the analysis and testing of a reallife power distribution network. The results obtained were validated with the IEEE-9 bus and IEEE-30 bus systems. The results obtained for the Campus distribution network were not only highly revealing but it also provided comparatively information (in respect of GS versus NR) as follows: number of iterations (i.e. 3 versus 177) , convergence time (i.e. 0.2457 versus 0.3276), power mismatch (0.017 MVAr versus 0.00 MVAr), system losses (i.e. 0.854 MW versus 0.855 MW), iteration tolerance (0.00001 versus 0.00) From this, the compared results indicated that the NR method converges faster with a least number of iterations irrespective of the number of the system buses while in the GS method, the number of iterations increases proportionally as the number of buses increases. Thus, it is evidently established that the NR method is very adequate for the analysis of large distribution networks.
- ItemOpen AccessOptimal load frequency control of two area power system(Faculty of Engineering, Ahmadu Bello University, Samaru-Zaria, Nigeria, 2019-08) Oluseyi, P.O.; Akinbulire, T.O.; Yellowe, K.M.; Babatunde, O.M.; Alayande, A.S.Modern power systems are operated under various constraints which are meant to ensure an appropriate delivery of service. Meanwhile power system faces imbalance in power generation and its consumption in which the higher the load consumption the lower the frequency of operation. The governor-turbine combination will then experience a devastating reduction in its frequency of operation due to disturbance that may lead to system collapse. Both governor and turbine are included in the model of this power system. The control objective is to regulate the frequency error, tie-line power error and area control error despite the presences of external load disturbance (0.01 pu) and system uncertainties. Various control policies were investigated using various combinations of system parameters on a platform of a series of combination of the PI, fuzzy logic and neuro-fuzzy controller with the power system for better stability. This could be found in the other approaches. The neuro-fuzzy based controller load frequency controller is simulated on this two-area interconnected nonlinear power system. To verify the performance of the various controllers, the data from a typical hydrothermal power grid was adapted for the study. From the simulation results; it was recorded the neuro-fuzzy controller enjoyed a settling time of 5 seconds while under the same operating condition the system stability is achieved at 12 seconds using the PI controller. This, thus, demonstrates the robustness of the neuro-fuzzy controller in contrast to the fuzzy logic and proportional-integral (PI) controllers. This thus shows that the neuro-fuzzy logic controller is superior to the other two considered in this work. Hence for a two-area network, the neuro fuzzy approach is recommended for the steady state operation of the system so as to ensure the dynamic stability of the network.