vfb基于补偿模糊神经网络的立式转鼓水果分级系统设计

提出了能无级调速的转鼓、搓动鼓设计和一种具有快速学习算法的补偿模糊神经网络水果分级系统，该系统在融合模糊理论和神经网络技术的基础上，增加了补偿模糊推理层，实现水果智能分级系统的补偿模糊推理和模糊设计参数的自动调整，通过与其它神经网络比较，验证了所提出的方法具有学习速度快、分级精度高的优点；不变的转鼓上滚子节距，克服了卧式滚子链链关节使用过程中被拉长而影响分级机适应范围的不足；CCD位于转鼓一侧的上升段上，缩短了CCD采样时段上水果沿垂直于采样方向移动的距离，使采样图像更清晰。

System design of the vertical drum-turning fruit sorter based on compensating fuzzy neural network

This article puts forward the design of non-polar regulation transferring drum,rubbing drum and a systematic structure to study compensating fuzzy neural network fruit grading,which can perform calculation fast.This system is on the basis of merging the fuzzy theory and neural network technology,which has increased compensating fuzzy deducing layer and realized the compensating fuzzy deducing and automatic adjustment of the fuzzy design parameter of intelligent fruit grading,through the comparison with other neural networks,it has verified the advantage of the method:high efficiency in studying,high precision in hierarchical and strong ability in generalization,the invariable sub pitch overcomes the deficiency of the horizontal chain joint of the roller chain during the course of being elongated which influences the hierarchical accommodation;CCD lies in the rising section of the transferring drum,which shortens the distance that the fruit moves along the vertical sampling direction and makes the picture clearer.