关键词: 油菜（Brassica napus L.）, 气象, 决策树, 支持向量机, 预测

Abstract: Aiming to construct a local rapeseed single yield forecasting model based on analysis of the influence of meteorological factors on rapeseed yield in Anhui region, the datas containing surface meteorological data of 78 observation sites in Anhui province from 1999 to 2018 and rapeseed single yield of 78 counties （cities, districts） from 2000 to 2018 were collected. The decision tree algorithm was adopted to analyze influence of meteorological factors on rapeseed output which finally screened out the main factors. A predictive model of rapeseed output based on support vector machine （SVM） was constructed. The results showed that the main meteorological factors which affected rapeseed yield were wetting index during maturity, average temperature during bud stage, wetting index at seedling stage and flowering stage. The root mean square error between predicted and measured value was 402 kg/hm², and fitting index was 0.72 which proved the prediction model behaves well on the whole.

Key words: rapeseed（Brassica napus L.）, meteorology, decision tree, support vector machine, prediction