基于模糊认知图的气象要素预测模型

Prediction Model of Meteorological Elements Based on Fuzzy Cognitive Map

根据气象要素相关性,构建气象要素预测模型,依据模糊认知图的推理机制实现对气象要素的预测,将对人类的疾病预防、环境污染等方面产生重要影响。该文将模糊认知图运用于气象要素中,提出一种基于模糊认知图的气象要素相关性分析及要素预测的模型构建方法。此模型综合考虑气象要素之间相互影响相互制约的关系,将气压、露点温度、湿度等九种不同的气象要素作为模糊认知图的概念节点,通过相关分析法将节点之间的相关程度数值化为各概念节点间的权重,得到初始关系矩阵,利用遗传算法对初始关系矩阵进行优化,优化后的关系矩阵和模糊认知图相结合对未来的气象要素进行预测。实验证明,该模型对气象要素的预测精度较高,实现了对气象要素预测的要求。但是实验中使用的气象数据规模有限,对模型的泛化能力有一定影响。总体来说该研究对推动气象要素预测具有积极的意义。

According to the correlation of meteorological elements,the prediction model of meteorological elements is constructed.The prediction of meteorological elements is realized by the reasoning mechanism of fuzzy cognitive map,which will have an important impact on human disease prevention and environmental pollution.In this paper,fuzzy cognitive map is applied to meteorological elements,and a model construction method based on fuzzy cognitive map is proposed for correlation analysis and element prediction of meteorological elements.The model comprehensively takes into the mutual influence and mutual restriction between meteorological elements account.Nine different meteorological elements,such as air pressure,dew point temperature and humidity,are regarded as the concept nodes of fuzzy cognitive map.The correlation degree between nodes is numerical converted into the weight among the conceptual nodes by correlation analysis method,and the initial relation matrix is obtained.The initial relation matrix is optimized by genetic algorithm.The optimized relation matrix and fuzzy cognitive map are combined to predict the future meteorological elements.The experiment shows that the model has high precision and meets the requirement of meteorological element prediction.However,the scale of meteorological data used in the experiment is limited,which has a certain influence on the generalization ability of the model.Generally speaking,this study has positive significance to promote the forecast of meteorological elements.