Process performance evaluation using evolutionary algorithm

Nowadays every business is using different quantitative measures and techniques to assess peiformance of their products! services. It is well known that different manufacturing processes very often manufacture products with quality characteristics that do not follow normal distribution. In such cases, fitting a known non-normal distribution to these quality characteristics would lead to erroneous results. Furthermore, there is always more than one characteristic Critical to Quality (CTQ) in the process outcomes and very often these quality characteristics are correlated with each other. In this paper, we assess peiformance of such a bivariate process data which is non-normal as well as correlated. We will use the geometric distance approach to reduce the dimension of the correlated non-normal bivariate data and then fit Burr distribution to the geometric distance variable. The optimal parameters of the fitted Burr distribution are estimated using Evolutionary Algorithm (EA). The results are compared with those using Simulated Annealing (SA) algorithm. The proportion of nonconformance (PNC) for process measurements is then obtained by using the fitted Burr distributions based on the two methods. The results based on both search algorithms are then compared with the exact proportion of nonconformance of the data. Finally, a case study using real data is presented.