• Login
    View Item 
    •   MKSU Digital Repository Home
    • Research and Publications
    • School of Engineering and Technology
    • School of Engineering and Technology
    • School of Engineering and Technology
    • View Item
    •   MKSU Digital Repository Home
    • Research and Publications
    • School of Engineering and Technology
    • School of Engineering and Technology
    • School of Engineering and Technology
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    24-Hours Load Forecasting Using a Hybrid of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) for Optimized Neural Network

    Thumbnail
    View/Open
    Full Text (1019.Kb)
    Date
    2016
    Author
    TabanInyasioSiriOgugu, Stephen
    Abungu, N.O
    Wekesa, Cyrus Wabuge
    Metadata
    Show full item record
    Abstract
    Short-term load forecasting (STLF) has emerged as one of the most important fields of study for efficient and reliable operation of power system. It plays a very significant role in the field of load flow analysis, contingency analysis, planning, scheduling and maintenance of power systems facilities, therefore, the system n costeffective is determine by accurate load forecast. Numerous researches have been done to improve accuracy of the conventional methods such as time series, regression analysis or ARMA and the use of artificial neural networks in load forecasting, ANN has shown lucrative results. But their training, with back-propagation algorithm or gradient algorithms is encounter long processing time, difficulty in selecting the optimal order of the components and trapping in local minima. These researches aimed at solving this problem by proposing a hybrid based on GA and PSO for train ANN and optimize the weights of ANN. The proposed method utilized the merit of GA ability for exploration of solution in space and PSO is widely well-known by social interchanges thinking ability. This help in reducing the search space for the algorithm thus reducing the iteration time. The proposed algorithm was tested using 24 hourly load data of different days i.e. Weekdays and weekends, the results obtained were compared to those results obtained by other researchers. It was observed that HGAPSO-ANN method has a better performance in terms of reducing and improving forecast error compared to Heuristic, PSO-ANN and GA-ANN methods. It was investigated that the method the lowest APE results is HGAPSO-ANN with an approximate minimum average error is 2.734% maximum average error 6.805%, therefore, a hybridized HGAPSO algorithm with ANN to help in reducing and improving forecast error. The suggested method was programmed under MATLAB 2016® software.
    URI
    http://ir.mksu.ac.ke/handle/123456780/2108
    Collections
    • School of Engineering and Technology [115]

    DSpace software copyright © 2002-2015  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    @mire NV
     

     

    Browse

    All of Digital RepositoryCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsBy Submit DateThis CollectionBy Issue DateAuthorsTitlesSubjectsBy Submit Date

    My Account

    LoginRegister

    DSpace software copyright © 2002-2015  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    @mire NV