Optimization Retrieval of Subtropical High Dynamic Forecast Model Based on Genetic Algorithm

Aiming at the difficult of establishing subtropical high dynamic model, the sum of the square difference between actual data and model output was regarded as target restriction, by introducing and applying the idea of dynamic model parameters retrieval and genetic algorithm (GA), a climate-meaning non-linear dynamic model of Pacific subtropical high was retrieved from the time series of 10-year averaged Pacific subtropical high characteristic indexes. The simulative forecast results of the retrieved dynamic model showed that for GA having the connotative advantages of global optimization and parallel calculation, the modeling method of dynamic forecast model by GA can accurately describe and simulate the subtropical high activity, and can be used for diagnosing and predicting such complicated weather system as Pacific subtropical high