机构地区: 山东农业大学信息科学与工程学院
出 处: 《数学的实践与认识》 2010年第6期92-98,共7页
摘 要: 对水文中长期预报模糊识别方法进一步研究,基于模糊环境下的目标函数,提出了具有主观监督因子和稳定系数的模糊识别预报模型.根据已知样本的最优模糊划分建立预报模型,利用已知样本的指标和样本的最优模糊划分计算预报模型的参数,给定模型的稳定系数,再通过调整主观监督因子对预报模型参数进行优化.径流中长期预报实例的模型检验平均相对误差为7.84%. The paper has further studied the algorithm of fuzzy recognition for mid and long term runoff forecasting,based on the new fuzzy objective function with subjective supervision factor and stability coefficient,the fuzzy recognition model for mid and long term runoff forecasting is presented.The optimization principle is that forecasting model was built according to known samples' best fuzzy classification,and the forecasting model's parameters are determined by using samples' index and best fuzzy classification,and the forecasting model's parameter are optimized through adjusting subjective supervision factor at given stability coefficient. The method presented in this paper is applied to runoff forecasting,the predicted result's mean relative error is 7.84%.