摘要
针对模糊神经网络学习算法计算量过大的问题,在预测模型设计中,提出了一种基于偏移率的模糊推理算法。此模糊推理算法与现有的模糊推理方法相比,有较少的计算量,且满足其模糊凸性和正规性。
Due to the computation of the fuzzy neural network learning algorithm oversized,a fuzzy reasoning algorithm based on moving rate is proposed in the forecast model design.This new method which compared with the existing fuzzy reasoning method has the few computation loads,and satisfies its fuzzy convexity and the normality.
出处
《机电产品开发与创新》
2011年第1期1-3,共3页
Development & Innovation of Machinery & Electrical Products
关键词
偏移率
模糊神经网络
模糊推理
预测模型
moving rate
fuzzy neural network
fuzzy reasoning
predicative
