New CBR adaptation method combining with problem–solution relational analysis for mechanical design Case based reasoning (CBR) methodology is proved to be a promising methodology on determining the parameter values of new mechanical product by adapting previously successful solutions to current problems. Compared with the sophisticated case retrieval technique, the case adaptation under K-nearest neighbor is still a bottleneck problem in CBR researches, which needs to be resolved urgently. According to the characteristics of parametric machinery design (PMD), i.e., less data and many parameters, this paper employs weighted mean (WM) as a basic model, and presents a new CBR adaptation method for PMD by integrating with problem–solution (PS) relational information. In our proposed adaptation method, prior to adapting the similar cases, the grey relational analysis (GRA) is utilized to investigate the PS relational information hidden in K retrieved cases, and the proposed method is called as GRA-WM. Different from classical WM method, the weighting factor of retrieved case for each solution element adaptation is calculated by multiplying similarity matrix (SM) and relational matrix (RM), and the adapted solution values of new mechanical product are subsequently obtained by calculating the weighted average of solution values of K similar cases. A case study on the power transformer design is given to prove the industrial applicability of GRA-WM. Moreover, the empirical comparisons between GRA-WM and other adaptation methods are carried out to validate its superiority. The empirical results indicate that GRA-WM can offer an acceptable adaptation proposal in application of CBR for mechanical design 35744
1. Introduction
Nowadays, the survival and success of mechanical companies largely depend on their ability to develop new mechanical products within a short period of time. One of the most widely used strategies for rapid machine development is the parametric machinery design (PMD) which has been known to reduce design time with minimal expense, and designers can make a comparatively reasonable decision by inputting some easy-to-assess parameters into the PMD system. However, PMD is a complex problem as there are massive parameters in the process of design and the model of mechanical product is complicated. So plenty of domain knowledge is required in traditional design which designs the product step by step according to design handbooks, and the design period is time-consuming. For many mechanical companies, it is widely accepted that their design practices rely heavily on past design experiences, instead of designing everything from scratch. Case based reasoning (CBR) methodology is built based on this concept, which is proved to be a promising methodology on determining the parameter values of new mechanical product by adapting previous successful solutions to current problem. So far, CBR has been employed to assist various machinery products design, e.g., micro-electromechanical device, test turntable, low power transformer, extrusion die bearing device, gear reducer and welding fixture.
Currently, the most existing CBR researches for mechanical design are predominantly characterized by a sophisticated case retrieval technique. For example, Li and Xie presented a new customer requirements-driven case retrieval method for fruit chute system design. Guo proposed an intelligent retrieval method by integrating ontology technology into CBR system to design new injection mould. Xie researched a retrieval method based on cost function to handle the missing values and unmatched features in hydro-generator design. Ref focused on the fuzzy case retrieval method to design new power transformer. Nevertheless, in the application of CBR, the old solutions of retrieved cases cannot be applied directly to the new problem, particularly in design domain, due to the complexity of problem requirements. Although some reuse-oriented retrieval methods have been put forward, these methods still focus on retrieving and evaluating similar designs at either the product level or the part level only. Thus, the contribution of case retrieval in CBR is limited as it is incapable of determining the final solution. From this point, the case adaptation is urgent for CBR application. For case adaptation, early CBR systems merely perform the retrieval-and-reuse task in the principle of 1-nearest neighbor namely, the solution values of most similar case is modified by decision-maker to fit the new problem, and this approach heavily depends on human subjective judgments. How to take advantage of K similar cases to generate more accurate solution for new problem, i.e., case adaptation under K-nearest neighbor, becomes a popular direction in CBR researches. In recent years, different adaptation models using statistic or intelligent methods have been proposed in the fields of environment, medicine, construction, finance and management. Both of them have advantages and disadvantages, and the related works will be discussed in the following section. According to the characteristic of less data and many parameters in PMD, this paper proposes a new adaptation method for mechanical design, and gives an example to investigate its feasibility. Moreover, we attempt to investigate whether or not the new method can produce higher adaptation performance than classical methods. CBR方法分析机械设计问题解决方案英文文献和中文翻译:http://www.youerw.com/fanyi/lunwen_33839.html