Electric Machines & Power Systems · 1995 · 21 citations · 9 references
Power EngineeringMachine LearningEngineeringPower System ApplicationsOperations ResearchSystems EngineeringAnn ModelsPower SystemsPower System AnalysisBack Propagation AlgorithmElectrical EngineeringIntelligent OptimizationDemand ForecastingEnergy ForecastingComputer EngineeringPower System OptimizationComputer ScienceEnergy PredictionPower NetworkIntelligent ForecastingSmart GridEnergy Management
ABSTRACT Artificial Neural Networks (ANN) are gaining popularity in various fields of engineering including electrical power systems due to their high computational rates and robustness.One of the ANN models extensively used for power system applications is the multilayer perceptron model based on back propagation algorithm. However,its training requires large number of input-output data sets which increases with system size and may become prohibitively large and time extensive. Moreover, the back propagation algorithm offers slow convergence with random initial weights. This paper presents a new approach to minimize the number of training patterns for ANN by using variable slope of the sigmoidal function for different test cases. In addition, the paper suggests the use of new functions for generating initial weights for training. The ANN models so developed have been tested to solve economic load dispatch (E.L.D.) problem on IEEE-14 bus test system and 89-bus Indian system. The proposed approach provides tremendous saving in the training time of ANN and provides fast and accurate results of E.L.D.
9