摘要： 以可拓检测为研究基础，提出第3类检测问题：待测物元转化问题，即新的检测环境下的可测物元与待测物元的互换问题。针对该问题的特殊性，以基于神经网络的检测模型为问题求解框架，可拓变换与关联函数作为遗传算法的设计准则，提出基于可拓遗传算法和神经网络的待测物元转化问题的解决方法。通过实际例子的应用验证了该方法的可行性。

关键词: 可拓检测, 待测物元转换问题, 可拓变换, 关联函数, 神经网络, 遗传算法

Abstract: This paper takes the extension detection as a foundation and puts forward the third detection problem which is measured matterelement exchange problem. It is detectable and measured which exchanges in the new detection environment. Considering the particularity of this problem, this paper sets the problemsolving framework based on the model of neural network, designs genetic algorithm on the basis of extension transformation and related function, and proposes a solution to this detection problem by extension genetic algorithm and neural network. It is verified by the application of examples.

Key words: extension detection, matterelement exchange problem, extension transformation, related function, neural network, genetic algorithm

中图分类号: 

• TP311.52