Analysis of Wheat Vulnerability to Frost Disaster Based on Hybrid Model of Fuzzy Neural Network

摘要

In China, frost is one of the major agro meteorological disasters. The analysis of the vulnerability of agricultural bearing body to the frost hazard has an important meaning on agricultural frost disaster risk assessment. In this paper, based on the theory of the natural disasters risk assessment and historical data of yield reduction; we analyzed the vulnerability of wheat to frost hazard in North China. With the hybrid model, we build a complex non-linear relationship of the yield reduction under different frost hazard intensities, and the wheat vulnerability to frost could be analyzed. The result shows that the hybrid model has fewer errors and a better effect than BP artificial neural network and linear regressive method. The result of the vulnerability curves of wheat to frost obtained by HM demonstrates the adaptability of the model in agricultural natural disaster risk assessment.