Abstract: Based on summarizing dynamic relationships between phosphorus and potassium fertilization and variety type, ecological factors and farming conditions from studying the literature and consulting experts, a rice knowledge model for design of suitable phosphorus and potassium fertilization under varied environments was developed through knowledge engineering and system modeling methods. The model can quantify the suitable total rates of phosphorus and potassium, ratios of basal to top dressing phosphorus and potassium under different environments and production systems. Case studies on the model with the data sets of typical varieties, soil fertility levels, farming systems and field comparative experiments indicated a good performance of the model system in applicability and decision making.

Key words: rice, phosphorus fertilization, potassiun fertilization, knowledge model

摘要： 在查阅和研究大量文献资料及咨询农学专家的基础上，通过总结、归纳和提炼水稻磷钾肥运筹与品种类型、生态环境和农作条件的动态关系，运用知识工程和系统建模方法，提出了具有时空适应性的水稻适宜磷钾肥运筹动态知识模型，可用于定量设计不同环境和生产条件下水稻品种的适宜磷钾肥用量及磷钾肥基、追肥比例。利用南京、常德两个代表性生态点的典型品种、不同土壤肥力、生产条件及大田对比试验对所建知识模型进行了实例分析和实践检验，结果表明模型预测结果与当地实际情况基本吻合，知识模型在磷钾肥运筹方面具有较好的适用性和决策性。

关键词: 水稻, 磷肥运筹, 钾肥运筹, 知识模型