基于不同温度和大麦种群不同生长阶段(软组织、叶片发育阶段等)的麦蚜种群实验数据,首先利用自适应神经模糊推理系统(ANFIS),分别建立高斯型、三角型和梯型隶属度函数的麦蚜种群内禀增长率的T-S初始模糊模型.然后在误差允许范围内,应用ANFIS具体训练流程对相应初始模型进行修正,并通过误差选择与原始数据拟合程度最好的模型,最后利用数值模拟验证结论的正确性.
Bases on experimental data of the wheat aphid population in different temperatures and different growth stages( soft tissue,leaf development stage etc),firstly,some initial models of intrinsic growth rate of aphid population with three membership functions are proposed by adaptive neural-network based fuzzy interference system( ANFIS),respectively. Secondly,for a given error,the corresponding models are modified using specific training processes of ANFIS,and the model which fits the date better than other models is chosen from the initial models above.Finally,the simulations are carried out to prove the result of the paper.