基于RS-TOPSIS的再制造曲轴毛坯质量评价方法

针对不确定条件下的再制造发动机曲轴质量等级划分问题,研究了基于粗糙集属性约简方法,并将逼近理想解的排序法运用到该问题中。选取主轴颈最大磨损量、连杆颈最大磨损量和主轴颈圆度等10个因素作为再制造曲轴的质量评价指标,利用粗糙集理论的离散和约简算法,将特征属性进行等级划分并计算出对应的特征属性权重,在约简结果基础上建立质量等级评价矩阵,通过计算评价对象的相对贴近度给出了质量分类区间。选取3×200组实测曲轴数据,建立RS-TOPSIS再制造曲轴质量等级评价模型,并对检测样本进行回检,误判率最大为2.1%。结果表明:所采用的RS-TOPSIS评价模型能够对再制造曲轴进行合理的等级评定,能够获得良好的判别效果,具有实际工程价值。

Quality Evaluation Method for Remanufactured Crankshaft Blank Based on RS-TOPSIS

The attribute reduction method based on rough set (RS)was investigated for the problem of quality grading for remanufactured engine crankshafts under conditions of uncertainty. The technique for order preference by similarity to ideal solution (TOPSIS) method was also used for the problem.Ten quality evaluation criteria including the maximum abrasion of the main journal crankshaft and the connecting rod and roundness of the main journal crankshaft were first selected. All feature attributes werethen classified and their corresponding weights were calculated by using discrete and reduction algorithm of rough set theory, and a matrix to evaluate quality classification was established based on these results.The quality classification interval was finally determined according to the relative closeness degree of evaluation objects.3×200 groups of data of measured crankshaft were selected to develop the RSTOPSIS quality grade evaluation model of remanufactured crankshaft. The test sample was checked to have 2.1% misjudgment rate. The results show that the RSTOPSIS method proposed can grade thequality of remanufactured crankshafts rationally and achieve good result, and thus would be useful in engineeringpractice.