基于动态模糊神经网络的产品成本估算

针对在产品方案设计阶段成本估算信息少且颗粒度大的问题，结合神经网络和模糊工程技术提出了动态模糊神经网络（DFNN），采用模糊推理的信息处理方法，在学习过程中隐层层数及维数根据规则不断变化，神经网络结构呈现动态.研究了动态模糊神经网络的学习过程、网络动态算法及模糊知识处理方法，建立了成本估算模型，并开发了基于动态模糊神经网络的成本估算软件，实现了利用产品方案设计信息自动进行成本估算.以挖掘机和液压油缸为例进行验证，结果表明该方法具有较强的信息处理能力，并提高了成本估算模型的柔性.

Product cost estimation based on dynamic fuzzy neural network

A new dynamic fuzzy neural network(DFNN) was proposed based on neural network and fuzzy engineering according to the mathematical expression of cost estimation.To resolve the problem of little information provided and the big grain degree of cost estimation in the product conception design process.DFNN is dynamic in the learning process due to the continual shift of the number of recessive layers and dimensions based on the fuzzy knowledge.After analyzing the learning process and fuzzy knowledge of DFNN,DFNN dynamic arithmetic methods of DFNN were presented including layer dynamic,dimension dynamics and function dynamics. DFNN was used to estabish the framework of cost estimation and the cost model for product conception design were established.The developed corresponding software based on DFNN was proved to has more powerful information dealing ability through examples of an excavator boom an oil crock.The DFNN also improves the flexibility of the cost estimation model.