基于结构模型的测试性设计与分析技术研究

Research on testability design and analysis technology based on structure model

为了解决因系统信息获取困难导致装备早期测试性设计和维修诊断工作难度增加的问题,提出了一种基于系统结构模型的测试性设计与分析方法。该方法借助图论的数学工具对系统进行数学建模,通过定量分析完成系统的模块划分;在此基础上,建立测试相关性矩阵,以最小测试代价为优化目标函数,应用Huffman信息编码方法生成系统的故障诊断树。与其他方法相比,该方法不仅结果更加优化,而且对系统的内部信息依赖相对较少,可以有效应用于设备的早期测试性设计以及使用过程中的维修诊断工作。

To solve the growing difficulties of early testability design and fault diagnosis caused by the difficulty to acquire information, a testability design and analysis method based on structure model is proposed. It builds up the mathematical models of equipments with the aid of the graph theory, accomplishes system modular division by means of quantitative analyzing. And then, the correlativity matrix based on structure model can be build up, it gains the fault diagnosis tree by selecting the least test cost as a optimal objective function and using Huffman information coding method at the same time. Comparing with other methods, this method has much more advantages, such as the results are more optimal and it less depends on internal information of equipments, so it can apply in the early testability design and maintenance work effectively.