基于VB环境的砂轮智能选择系统开发

VB Based Intelligent System Development for Selection of Grinding Wheels

提出用模糊自适应调整BP算法建立磨削条件与砂轮特征参数之间的关系模型;并依据磨削数据手册中的数据及实验结论对模型进行训练,训练好的神经网络可实现砂轮型号的智能推荐;测试结果显示模糊自适应BP网络比原BP网络收敛速度快、识别准确率高·该方法具有不依赖经验工人,无需要建立规则,利用最新的研究结果可随时训练网络,具有很好的扩展性等优点·利用VB和Matlab语言相结合的方法开发本系统,开发过程方便快捷,并具友好的用户界面·

?The fuzzy adaptive BP neural network was proposed to build the relationship model between grinding conditions and main character parameters of a grinding wheel. The model was trained by the samples from 'Grinding Data Book' and experiment data. The trained networks model can be used to recommend suitable grinding wheel type for the given grinding conditions. The testing results indicate that the fuzzy adaptive network has faster convergence speed and higher accuracy than the standard BP network does. This method is independent on skilled workers, and has good expansibility. It need no rules, and can retrain the network by new research results.