A novel Local BP Neural Network model and application in parameter identification of power system

The traditional mathematical modeling is nonrepresentational and it is hard for understanding. In oder to model the real system in a intuitive method, a novel Local BP Neural Network (LBPNN) model has been proposed to imitate arbitrary feed-forward topologies of networks and the weights' training algorithm—constrained stochastic gradient descent (CSGD) is also introduced in this paper. The network model could be used to approach functions which could improve the training speed and reduce the training complexity. With the LBPNN model and the CSGD training algorithm, network's parameter of weight could be identified within constrains. The training algorithm's robustness and effectiveness are verified in the LBPNN. Finally, the LBPNN model is used to map the fuzzy petri net(FPN) of a power system and the parameters of weights in the FPN are identified by training the LBPNN model with the CSGD algorithm.