摘要
针对施肥过程中普遍存在非线性和不确定性问题,对多种模糊神经网络进行分析比较,提出了适合施肥决策系统的FMLP模糊神经网络结构,采用FBP学习算法实现网络结构优化。通过对水稻在同一农田重复进行种植实验,结果表明所建系统可以提供最佳土壤施肥方案。
The paper raises the FMLP Fussy-Neural Network to suitable fertilizer decision facing universal problems of nonlinearity and uncertain comparing kinds of FNN during applying fertilizer. The network structure is optimized by adopting FBP learning algorithm. The result demonstrates that the system can provide optimum scheme of fertilization by repetition experiment of planting rice in the same farmland.
出处
《农机化研究》
北大核心
2009年第3期175-176,179,共3页
Journal of Agricultural Mechanization Research
基金
黑龙江省自然科学基金项目资助(F0326)
东北农业大学博士启动资金项目(2008)
