【摘要】 探索机器学习在轮胎配方设计中的应用。结果表明：可以利用数据挖掘的方法进行配方设计中单个性能的优化，对其他重点指标，可以增加1个分类指标进行控制，从而达到性能优化的目标；可以利用聚类原则进行配方优化，但该原则仅代表取得较好性能的概率较大，并非选取的优化点肯定都好。更多还原

【Abstract】 The application of machine learning in tire formula design was explored. The results indicated that data mining methods could be used to optimize individual performance indicators in formula design,and one additional classification indicator could be introduced to control other key indicators,thereby achieving the goal of performance optimization. Formulation optimization could also be carried out based on clustering principles,but such principles only indicated a higher probability of achieving better performance, rather than guaranteeing that all selected optimization points would yield ideal results.更多还原