为了在考虑架构约束条件的产品模块化过程中直观地识别出模块, 提出一种基于可视化对角矩阵的方法.首先通过包含架构约束条件的遗传算法自动产生一组优化的模块划分方案; 其次根据优化解集构建出成组可能性矩阵(GLM)并对其对角化, 得到对角 GLM(DGLM); 再将 DGLM 的非对角单元根据可能性值进行着色; 最后通过DGLM辨识出系统的典型结构和各种潜在的模块. 以磁共振成像设备中的注射器为实例, 验证了该方法的有效性.
In order to intuitively identify modules in the product modularization process with considering the architectural partitioning constraints, a visual diagonalized matrix-based method was proposed. First, a ge-netic algorithm incorporating architectural constraints was used to generate automatically a set of optimized module partition solutions. Then, a grouping likelihood matrix(GLM) was obtained using these optimized solutions and was diagonalized to form a diagonalized GLM (DGLM). Next, off-diagonal cells of the DGLM were color-encoded according to their likelihood values. Finally, the typical system structures were dis-played in the visual DGLM, and potential modules were identified. A case study of designing a MRI machine injector was carried out to verify the effectiveness of the proposed method.