Application Research On Reliability Growth In The Process Of Product Design

With the rapid development of modern technology, the emerging of new productsand the accelerated replacement cycle of products, that require the better reliability ofnew products. The products of high reliability are more safe and practical. And they cancreate large economic benefits. Only the companies with high-reliability products cansurvive in the market of increasing fierce competition.The paper reviews the current situation and problems of reliability development athome and abroad. The method, test and model of reliability growth are researched in thedevelopment process of new products. The plan, evaluation, test program andmathematical models of reliability growth (the classic and Bayesian model) are analyzedin the product design process. Finally, the characteristics of classic model (Duane modeland AMSAA model) and Bayesian model, and the application for different size samplesare analyzed. Therefore, the theoretical results of reliability growth are sorted out andthe new applications of reliability growth are explored in the design process of newproducts.The paper discusses the analysis methods of reliability growth in the product designprocess. And they include FTA, FMECA and Monte-Carlo method. The characteristicsand application scope of various reliability analysis methods are analyzed. In the productdesign process, a new concept to improve the reliability growth of products has beenproposed.The paper presents an analysis method of combining a large and small sample forthe reliability of new products in the product design process. For the productcharacteristics of a large sample, more mature classic reliability growth analysis mode isapplied. And for the product characteristics of a small sample, Bayesian reliabilitygrowth analysis mode is applied. For a small sample system, the fault simulation isobtained in the system design process with system simulation. Combining the actualsystem design, the potential failure modes of the small sample system are analyzed.Through the historical experience, the level of system reliability can be improved.In this paper, the series-parallel model of the size samples of new product isanalyzed and the Bayesian network model of products is established. The effects of thesubsystem reliability of size samples on the reliability of new products are analyzed. Finally, the correct analysis and evaluation of the reliability of new products can beobtained.In order to achieve the above reliability growth program, the failure mode andsolutions of historical product should be researched using the data mining methods andthe similarity principle. Based on the knowledge-base of reliability growth, theexpression forms and search technology of knowledge-base are analyzed, and thedatabase system of reliability growth analysis can be researched and developed. Anyfault information can be saved in the product design process and the reference value canbe provided for the development of new products.Finally, on the basis of the above studies, the large servo presses and aluminumalloy castings are selected for researching and application verification.