关键词: 设施黄瓜, 生长模型, 主成分分析, 岭回归, 多重共线性分析

Abstract:  At present, cucumber planting management standards are low and environmental factors are difficult to be controlled in an accurate way in the process of crop growth. To address the two problems, the paper took cucumbers planted during autumn and winter in Lingcheng District, Dezhou City, and made use of PCA (principal component analysis), multicollinearity analysis, and ridge regression algorithm to build the facility cucumber growth model based on major environmental factors. The result showed that firstly, in the growing indicators such as plant height, stem diameter, number of leaves and leaf area, leaf area load value in the seedling stage, and the initial flowering stage was 0.994 and 0.980, which could show that the leaf area can obviously reflect growth conditions at the seedling stage and initial flowering stage.Secondly, the main environmental factors that affect leaf area change at the seedling stage were illumination, temperature, and humidity. The main environmental factors that affect the leaf area change at the initial flowering stage were temperature and light. Thirdly, for the facility cucumber growth model in different periods based on the ridge regression, the model determination coefficient was greater than 90%. The growth model built through the research has stronger pertinence, which has provided better direction for scientific selection of growth indicators of cucumbers at the seedling stage and initial flowering stage, and given a reference for facility cucumber planters to adjust and control the production environment.

Key words: facility cucumber, growth model, principal component analysis, ridge regression, multicollinearity analysis