Forecasting Method For Short-term Load Based On Quantum Neural Network

Short-term load forecasting is an important routine for power system. How to improve the forecasting accuracy is the emphasis on the study of short-term load forecasting. This paper based on analyzing quantum computing and artificial neural network combined the specialty of the real load proposes one short-term load forecasting method with quantum neural network (QNN). The main researches are as follows: Firstly, the text gives a summarization for present method of short-term load forecasting and analyzes the interior features and exterior influence factors of load. The load forecasting models and evaluation function are proposed. The input feature vectors are disposed.Secondly, a three-step quantum neural network model based on quantum bit neural unit is designed for short-term load in this paper. The quantum gate as quantum neural network's activation function implements quantum neural computation. The model using complex number and the real and imaginary working at the same time reflect quantum parallel characteristic. The weight and threshold of output layer carrying out non-linear transform achieve nonlinear mapping function. Using Genetic algorithm optimizes the network's initial parameter. The paper uses MATLAB program to simulate and the result shows that the convergence rate and functional capacity and predicting precision of this model are prior to those of the normal three-step BP network.Thirdly, make the MATLAB simulate program import to load forecasting software and evaluate the method. The method provides a viability solution for forecasting of power system.