Short-term tidal level forecasting based on self-adapting PSO-BP neural network model

Real time tidal level prediction is essential for management of human activates in coastal and marine areas. However, prediction model with static structure can not represent the variations caused by time-varying factors such as weather condition and river fluctuation. The traditional harmonious analysis is not able to achieve high prediction accuracy in prediction of tidal level with nonlinear and no-stationary characteristics. In order to resolve the problem, we propose a self-adapting particle swarm optimization (SAPSO) algorithm to optimize the back propagation (BP) neural network model. The proposed model is referred to as SAPSO-BP model which employs PSO to adjust control parameters of BP network. To validate the effectiveness and efficiency of the proposed SAPSO-BP model in tidal level forecasting, real-measured tidal level data of Port of Honolulu was chosen as the test database and the result shows that short-term tidal prediction coincides well with the measured tidal level.