Traditional product data mining methods are mainly focused on the static data. Performance requirements are generally met as possible by finding some cases and changing their structures. However, when one is satisfied with the structures changed, the other effects are not taken into account by analyzing the correlations; that is, design conflicts are not identified and resolved. An approach to resolving the conflict problems is proposed based on propagation analysis in Extension Theory. Firstly, the extension distance is improved to better fit evaluating the similarity among cases, then, a case retrieval method is developed. Secondly, the transformations that can be made on selected cases are formulated by understanding the conflict natures in the different performance requirements, which leads to the extension transformation strategy development for coordinating conflicts using propagation analysis. Thirdly, the effects and levels of propagation are determined by analyzing the performance values before and after the transformations, thus the co-existing conflict coordination strategy of multiple performances is developed. The method has been implemented in a working prototype system for supporting decision-making. And it has been demonstrated the feasible and effective through resolving the conflicts of noise, exhaust, weight and intake pressure for the screw air compressor performance design.