Non iterative-state estimation based neural network for short term load forecasting of distribution systems

In this paper, a new non-iterative state estimation based neural network is proposed for solving short term load forecasting of distribution systems. In this approach, the weights between the layers of neural network have been estimated using the weighted least square state estimation (WLSSE) technique without any iterative approach. The WLSSE technique could offer well established weights by accounting noise associated with the input and output data. This approach is very fast compared to conventional back propagation technique. The proposed method can predict day-ahead loads of the distribution system. The effectiveness of the method is examined on practical NPDCL Warangal, Indian 132/33 kV substation distribution system. The method can provide more accurate results than the back propagation neural networks. The test results are compared with that of the results obtained from conventional ANN method.