Application of ARIMA-GRNN Model in Dam Safety Monitoring
Tengfei Bao · Water Resources and Power · 2013
To improve the prediction accuracy of the dam safety monitoring data,the autoregressive integrated moving average model(ARIMA) and the generalized regression neural network(GRNN) are combined to build the ARIMA-GRNN prediction model.With the pre-measured and ARIMA fitting values as the input and the late-measured values as the output,the best prediction model of GRNN network is established by optimizing the smooth factor which is based on the minimum mean squared error between late-measured and the predictive values.Entropy weight method and standard deviation method are used to evaluate indexes of each model comprehensively.The results show that the accuracy of the ARIMA-GRNN model is obvious higher than that of the ARIMA model,which means the feasibility of the ARIMA-GRNN model for application to the dam safety monitoring.