The application of the fractal and chaotic forecasting model in dam safety monitoring

The observed data sequence in dam safety monitoring is usually composed of the certainty component, chaotic component and random component. Though the statistical model can make good prediction, it can not reflect the chaotic component of the effect variables. Many environmental factors must be determined before the statistical model could be used. ne precision of the model will be affected if the environmental factors are not determined properly or the error of the environmental observed data is sort of large. Through the analysis of the certainty component, chaotic component and random component of the observed data sequence, fractal forecasting model and chaotic forecasting model are respectively established for the prediction of the certainty component and chaotic component based on fractal and chaotic theories, and the hybrid forecast model is established by the superposition of these two models. The application of the hybrid model indicates that it could get the same or even better forecasting result than the traditional statistical model and it avoids the use of environmental factors.